International Journal of Medical Engineering and Informatics (78 papers in press)
Development of a Serious Game to Enhance Assistive Rehabilitation
by Mauro Bellone, Andrea Martini, Francesco Argese, Piero Cirillo, Italo Spada, Antonio Cerasa
Abstract: The aim of this study is to investigate novel assistive technologies
based on serious gaming for the assessment of postural control and motor
rehabilitation. Previous research already demonstrated that rehabilitation and
physical activities can improve the quality of life of patients, and virtual reality
applications may act as good additional companions during the therapeutic
sessions. Indeed, workout routines supported by serious gaming encourage
patients to train harder, bringing the therapy towards a pleasant game into a
virtual environment. During the game, sensors track the movements of the
players and transfer the data to software components that record a database
from which patients progress can be determined. The acquired measures are
interpreted in the form of new biomarkers, which enable the assessment of
postural control. Such biomarkers are based on a probabilistic approach and
show the capability to discriminate between well-performed exercises and incorrect movements. A long-term experimentation on a specific exercise is
proposed, showing a considerable improvement, ranging from 5\% to 30\%, in
the performance of patients.
Keywords: rehabilitation; virtual reality; motor inabilities; visual tracking;
An Evaluation of Released Mobile Health Apps in Popular Iranian App Stores
by Hamid Naderi, Kobra Etminani
Abstract: The rapid growth in development and using mobile health applications had significant potential in the field of health. The study objective was to describe apps in the most popular Iranian Android and IOS app stores. Detail information of apps were extracted, categorized by subject and descriptive analyses were done. 3331 android and 277 IOS apps were remained for study. 88% of popular android mHealth apps were free on the other hand 70% of popular IOS apps were paid. The average price of android and IOS paid apps were $0.37 and $6.55. The average of rating value for all android subcategories (total average 4.33) is higher than IOS subcategories (total average 3.58). Our study confirmed fitness was the most popular topic. The average of rating value for all android subcategories was higher than IOS subcategories. It seems large number of mHealth app users preferred using applications which can update their content continuously.
Keywords: mHealth; Iranian app stores; Mobile applications; Descriptive analysis.
Digital Watermarking in Medical Imaging: A Review
by Hana Ouazzane, Hela Mahersia, Kamel Hamrouni
Abstract: This paper presents considerations on the digital watermarking potential to fit within the medical imaging field. We aim to identify watermarking applications and main strategies within different approaches through previous works. An analysis and evaluation of previous literature quality is also established. We conclude that digital watermarking has considerable purposes in the medical imaging field, it joins and complements existing security measures and management tools for medical images.
Keywords: digital watermarking; fragile/semi-fragile watermarking; robust watermarking; lossless watermarking; medical imaging; content security; e-health; medical information systems.
An Innovative Solution for prevention of reverse flow of blood in IV set (Intravenous set) using Embedded Systems
by Abhishek SN, Ikram Shah, K.V. Shriram
Abstract: The demand for medicine and doctors is never ending and it is really evergreen as our society has been revolutionized and the life expectancy rate has increased. Everyone starting from a six year old child to 90 year old veteran is aware of all kind of diseases exists as of now, specialization of doctors, which doctor to approach when we have a particular problem and so on. Thus, a common wish prevailing in everyone irrespective of age is that to live long and stay in pink of health. Adding to this point, millions and millions of innovations are being made in the medicinal field every decade to enhance the average life expectancy rate. There are various sectors of engineering that are involved in creating and/or maintaining various medical instruments. Say, from the smallest equipment like syringe to the largest one like CT scan machine have major contributions from the engineering sectors. IV Set (Intravenous set) is one among those substantial inventions in the medicinal field; it is actually used for transferring blood or saline from the IV/fluid bag to the patients body at a speed and pressure that the human body can accept. There is a notable threat associated with using the IV set which most of the doctors and care takers are aware of; When the fluid in the IV bag runs out (gets emptied) the patients blood would start flowing through the IV tube in the reverse direction towards the empty IV bag as the pressure of blood in the human body is more than the pressure experienced by the empty IV bag (diffusion), thus leading to blood loss and pain, at times. Thus the caretakers must be vigilant enough to replace the fluid bag at an appropriate time to avoid blood loss. The innovation proposed here is a solution that would prevent such blood losses and doesnt require manual monitoring of the IV bag thereby stopping the reverse blood flow well in time. This not only prevents blood loss but also reduces the stress for the care takers and pain for the patients. The product is tested and is costing less than 100 USD and is one-time investment for the hospitals.
Keywords: Intravenous set; Blood loss; intelligent IV Set; Hospitals; Medicine; Saline; Intelligent medical support; Patient monitor;.
Identification of Important Biomarkers for Detection of Chronic Kidney Disease using Feature Selection and Classification Algorithms
by E. Sivasankar, Pradeep R, Sivanandham S
Abstract: Health informatics plays a critical role to discover, innovate and reform the healthcare system towards betterment by fostering a collaboration between patients' data and medical practices. Development of easy, accurate and convenient methods for detection of diseases is a growing field of study in health informatics because of its easy scalability and affordability. In this paper, we have tried to identify important biomarkers for the detection of Chronic Kidney Diseases. We have used filter and wrapper based feature selection techniques and four different classification methods to analyse and interpret data to help us diagnose Chronic Kidney Disease faster. We have reported the top attributes and their predictive accuracy in detecting CKD and Support Vector Machine with Chi-squared feature selection method provided us with the best accuracy.
Keywords: Chronic kidney disease; CKD; Feature selection; Classification; Detection predictive model; Biomarkers for CKD; Risk factors; Machine learning; glomerular filtration rate; GFR; Nephrology.
Discrimination of signals phonocardiograms by using SNR report
by Mostefa Meriem, Hamza Cherif Lotfi, Debbal Sidi Mohammed El Amine
Abstract: The phonocardiographic PCG signal presents the acoustic recording of heart sounds; witch may reveal significant information that the human ear cannot. The current study develops methods for classification of signals using wavelet transform PCG, in Particular, the discrete wavelet transforms (DWT). In every case of PCG signal, different degree of noise will be gradually added and Ratio Signal to Noise (SNR) is Calculated. The intensity of the (SNR) is the basic parameter used in the classification of signals PCG.
Keywords: Phonocardiogram; Discrete wavelet transforms; SNR; Sound; ER.
An inventive and innovative system to detect fall of old aged persons A novel attempt with IoT, Sensors and Data Analytics to prevent the post-fall effects.
by Juluru Anudeep, Gudimetla Kowshik, Popuri Vamsi Krishna, Ikram Shah, K.V. Shriram
Abstract: According to the statistics of NCOA (National Council on Aging) these days rate of deaths in the old aged people has reached a critical state that for every 11 seconds an older adult is being treated for a fall and for every 19 minutes a death is taking place. Most of the old aged people are suffering from the back pain, joint pain, knee pain and they mostly are the ones who arent able to walk properly and are mostly constrained to bed. Statistics by north eastern Ohio state reveals that, the rate of emergency room treatment of fall-related injuries in persons aged 75 and older approached 80 per 1,000 per year in women and 60 per 1,000 per year in men. Falls are indications to some critical problems like immobility, fragility of bones and chronic health impairment. Falls in elders vary from that of the falls in kids, as the healing power in the old people slackens with their age. Falls are going to play a crucial role in the health line of the old people. Due to the falls there will even be a huge impact on the heart. Old people with heart diseases when falls suddenly, there is risk of being affected by Tachyarrhythmia which is typically a heart disorder in which the heartbeat raises to an abnormal rate of more than 100 beats per minute which is approximal to an cardiac arrest. So, in order to reduce these type of risks in elderly people , We present a frugal and affordable system that could monitor the motion of the old people and can detect their fall. Detection is not done before the fall or when person remains in rest state, it detects immediately after fall and alerts the concerned persons to do the necessary action to save the person. We use IoT, Sensors and Data analytics to build this system. The system is tested for its functioning and is fool proof.
Keywords: Fall detection; Old age support; IoT; IoT for medicine; Old aged tracking system; Gyroscope;.
Pragmatic Realities on Brain Imaging Techniques and Image Fusion for Alzheimers Disease
by P.S.Jagadeesh Kumar, Yang Yung, Mingmin Pan
Abstract: Alzheimers Disease is the absolute generous of age-related neurodegenerative syndrome. It specifies substantial brain atrophy, amnesia and major neuro-logic disintegration. Brain imaging techniques has been progressively employed for medical explanation and variant diagnosis, and to afford understanding into the paraphernalia on functional and structural measurements of the brain, sketches of dimensional distribution besides their normal antiquity and progression over time. This paper makes an exertion in performing a practical virtuosity on brain imaging techniques and image fusion for the diagnosis of alzheimers Disease. Several brain imaging techniques like computerized tomography, single-photon emission computed tomography, magnetic resonance spectroscopy, positron emission tomography, magnetoencephalography, magnetic resonance imaging diffusion tensor imaging was evaluated for alzheimers disease based on their degree of confidence, quality, volumetry, availability, cost and limitations.
Keywords: Alzheimer’s Disease; Brain Imaging Techniques; Image Fusion; Performancer Evaluation; Neurodegenerative Syndrome.
An Artificial Intelligence approach for the recognition of early stages of - ECZEMA
by Dhananjay Kalbande, Rohit Naik, Janvi Jatakia, Uday Khopkar
Abstract: The rural population of India suffers from various medical ailments and due to the lack of medical facilities and practitioners, enough support is not available. Medical help might come late and the problem might have been aggravated. With the increasing awareness about artificial intelligence (AI), it is possible to solve these problems using technology. The research aims at detecting an early stage of the skin disease eczema, when the affected part of the human body is captured through a smart phone and approximate symptoms are provided by the medical practitioner. It uses artificial intelligence algorithms like convolutional neural networks and support vector machines algorithm for classifying the images, and back propagation algorithm for training a model based on the symptoms. Around 50 clinical photographs of eczema acquired from KEM Hospital, Mumbai to train the classifier and then different images of eczema were tested with an accuracy of greater than 85%.
Keywords: artificial intelligence; back propagation network; BPN; convolutional neural network; support vector machine.
Transmission and Archiving of Reduced MRI Medical Images
by Hedi AMRI, Med Karim ABDMOULEH, Ali KHALFALLAH, Jean-Christophe LAPAYRE, Med Salim BOUHLEL
Abstract: Medical images are often large. To minimize their file sizes, different compression standards like JPEG, JPEG 2000 and TIFF have been developed, but they can affect the image quality in case of lossy compression. In this paper, we propose a compression method based on image resizing that can reduce the file size to a quarter before its transmission or archiving. When it is received or before its display, the reduced image is enlarged to ensure better visual comfort. This approach is called REPro. In this context, we have used three image reduction techniques namely Square-Square Mesh Decimation, Square-Square filtered Mesh Decimation, the Square-Staggered-Square filtered Mesh Decimation. In addition, we have utilized four enlargement techniques namely Zero padding, Nearest Neighbor interpolation, Cubic Spline interpolation and B-interpolation to resize MRI images. The experimental results prove that the combination of the Square-Square Mesh Decimation with the B-Spline interpolation ensures the minimum distortion of the original image.
Keywords: Telemedicine; Archiving; Transmission; Medical images; image Reduction; image Expansion.
Breast Cancer Detection in Mammogram image with Segmentation of Tumor Region
by Vikramathithan A C, Dandinashivara Revanna Shashikumar
Abstract: In our proposed breast cancer malignant detection study are performed with the aid of fuzzy min max neural network technique. Majority of womens are affected in this breast cancer at a early stage the mammogram images are mostly play in a vital role. Initially the input mammogram image smoothened with the aid of adaptive median filer from that smoothened image we are segmenting tissues with the aid of Histon based Fuzzy c-means clustering. We are extracting features from that segmented image the features are statistical and semantic features. Then we can identify the malignant region with the aid of these features. The segmented region is maligned or benign using an optimal fuzzy min max neural network with gray wolf optimization algorithm with the aid of these we will identify a breast cancer region.
Keywords: Breast Cancer; Mammogram; Fuzzy min max; gray wolf; optimization; segmentation.
Radial Basis Function Neural Network with Genetic Algorithm for Discrimination of Recombination Hotspots in Saccharomyces Cerevisiae
by Ashok Dwivedi
Abstract: Recombination influences the evolution of Saccharomyces Cerevisiae. Genomic regions where
recombinations occurs are known as recombination hotspots. There are two kind of hotspots for recombination. The spots where recombination occurs more frequently are called recombination hot spots and regions where recombination occurs less frequently are known as cold spots. In this work, we have formulated methods based on neural network models for the classification of these hot and cold recombination spots on the basis of compositional features of nucleotide sequences. These models were validated using tenfold cross validation technique. The classification accuracy of 83%, 82%, and 78 % were achieved using radial basis function neural network with genetic algorithm, radial basis function neural network and multilayer perceptorn models respectively. Moreover, the performance of these model were evaluated on differenct classification measurements. Furthermore, results indicate that redial basis function neural network with genetic algorithm gives best result.
Keywords: : Artificial Neural Network; Radial Basis Function Neural Network; Genetic Algorithm; Multi-Layer Perceptron; Classification; Machine Learning; Evolution.
Analysis of salivary components as non-invasive biomarkers for monitoring chronic kidney disease
by Navaneeth Bhaskar, Suchetha M
Abstract: Saliva, a valuable source of biochemical information, is a potential diagnostic substance that helps to identify many diseases. Studies have revealed that saliva tests help identify many diseases. Saliva test has excellent advantages over blood test as the former can be collected non-invasively using simple equipment. This paper explores how salivary components can be used as diagnostic tool to identify chronic kidney disease (CKD). Experimental analysis was conducted to assess the levels of salivary components in whole saliva of CKD patients in contrast with healthy people. Urea and creatinine are the most accepted biomarkers of CKD. The correlation between creatinine and urea levels in human saliva and blood were analysed. Unstimulated saliva flow rate and pH levels were also monitored in this study. The results obtained from this study give concrete evidence that there is a positive correlation between creatinine and urea levels in blood and saliva. From the derived regression line equations, serum urea and creatinine values can be predicted from salivary values. Receiver operating characteristics (ROC) performance analysis was performed and area under the curve (AUC) of 0.95 and 0.89 was obtained for salivary creatinine and urea.
Keywords: saliva; non-invasive; electrolytes; urea; creatinine; chronic kidney disease; CKD.
Time-Frequency Analysis based Method for Application of Infant Cry Classification
by J. Saraswathy, M. Hariharan, Wan Khairunizam, J. Sarojini, Sazali Yaacob
Abstract: This paper proposes a new investigation of time-frequency (t-f) based signal processing technique using wavelet packet spectrum (wpspectrum) for classification of newborn cry signals. T-f approaches can analyze the non stationary signals comprehensively in joint t-f domain compared to the time or frequency domain techniques. This study was initialized with the extraction of a cluster of t-f features from the generated t-f matrix of recorded cry signals using wpspectrum by extending time-domain and frequency-domain features to the joint (t-f) domain. In accordance, conventional features such as Mel-frequency cepstral coefficients (MFCCs) and Linear prediction coefficients (LPCs) were also extracted in order to compare the performance of the suggested t-f approach. Probabilistic neural network (PNN) and general regression neural network (GRNN) were used to evaluate the efficacy of the extracted feature vectors. The proposed methodology was implemented to classify different sets of infant cry signals cry including binary and multiclass problems. Best empirical result of above 99 % was reported and revealed the good potential of t-f methods in context of infant cry classification.
Keywords: Keywords: Infant cry; Time-frequency analysis; Feature extraction; Classification.
Improving ECG Signal Denoising using Wavelet Transform for the Prediction of Malignant Arrhythmias
by Agostino Giorgio, Cataldo Guaragnella, Domenico Andrea Giliberti
Abstract: Objective: This paper deals with the accuracy of algorithms for the detection of cardiac Ventricular Late Potentials (VLP). The presence of VLP in an electrocardiographic signal (ECG) is associated with possible sudden cardiac death, because of malignant arrhythmias. VLP detection is strongly influenced by signal noise, thus the ECG needs to be denoised before VLP detection. The objective of this paper is to define a denoising algorithm improving the VLPs detection in ECG signal, and to describe its hardware implementation on a Field Programmable Gate Array device (FPGA). Methods: The method described uses wavelet denoising, implemented as subband coding. The drawbacks of this method are heavy linear distortions undergone by the analyzed signal. This disadvantage is overcome by using an equalization filter properly designed by the authors for canceling the introduced distortions. Results: The algorithm (equalizer filter + wavelet denoising) has been firstly implemented and successfully verified using MATLAB. Then, it has been implemented as programmable hardware on Alteras FPGA. The synthesized hardware has been verified on the evaluation board DE1-SoC, mounting a Cyclone V 5CSEMA5F31C6 FPGA chip. Conclusions: On board processed results and theoretical results are consistent, validating the effectiveness of the algorithm and of the designed hardware. Significance: Results show that the algorithm accuracy and its capability to be implemented as programmable hardware also could be used for upgrading ECG devices reliability in the field of heart diseases prevention.
Keywords: Wavelet Transforms; Signal Detection; Biomedical Electronics; Denoising; FPGA.
Malignant Melanoma Detection using Multi Layer Preceptron with Visually Imperceptible Features and PCA Components from MED-NODE Dataset
by Soumen Mukherjee, Arunabha Adhikari, Madhusudan Roy
Abstract: In this paper, a scheme is worked out for classification of images belonging to malignant melanoma and nevus class by multi layer neural network architecture with different trainings and cost functions. Total 1875 shape, colour and texture features are extracted from 170 images from MED-NODE dataset. With the total 1875 features an accuracy of 82.05% is achieved. Feature ranking algorithm ReliefF is used for ranking these features. MLP is run with varying number of best ranked features. With 10 best features an accuracy of 83.33%, sensitivity of 86.77% and specificity of 72.78% are achieved with 3 fold cross-validation. Effect of pre-processing the features with Principal Component Analysis is explored and found that the optimal number of principal components is 25, which yields a maximum accuracy of 87.18% which is much higher than the previously reported accuracy level with this dataset.
Keywords: ABCD rule; GLCM; GLRLM; ReliefF; Principal Component Analysis; Autocorrelation; Cross Entropy; t-test.
A REVIEW ON AUTOMATIC IDENTIFICATION OF FOVEA IN RETINAL FUNDUS IMAGES
by Rajesh I S, Bharathi M A, Bharati M. Reshmi
Abstract: Identification of retinal diseases is a very significant area of ophthalmology. Regular procedures are extremely specific, which rely on manual observation and highly prone to error. Hence, it is extremely fundamental to set up an automatic system for screening of vision-threatening diseases like Diabetic Retinopathy (DR) and Diabetic Maculopathy (DM). Patients who are suffering from DR are at high risk to have DM which may lead to blindness, if not detected and treated appropriately at the appropriate time. Automatic analysis of retinal images requires knowledge and the properties of anatomical structures and retinal lesions. Thus, locating fovea plays a vital role in the analysis of retinal images. In recent times image processing has become a very effective tool for the detection and analysis of abnormalities in retinal images. This survey paper depicts the fundamental terminology related to automatic detection of macula and fovea. The literature review of various methods used for finding fovea in retinal fundus images is discussed. Detection issues involved in fovea are also discussed in this paper.
Keywords: Diabetic Retinopathy (DR); Diabetic Maculopathy (DM); Optic Disc (OD); Region of Interest (ROI); Blood Vessels (BVs).
Enhancement and segmentation of histopathological images of cancer using dynamic stochastic resonance
by Anuranjeeta Anuranjeeta, Munendra Singh, K.K. Shukla, Neeraj Sharma, Shiru Sharma
Abstract: Pathologists face difficulty in cell image detection as uneven dye causes the low contrast and inhomogeneity. The proposed Discrete Cosine Transform (DCT) based Dynamic Stochastic Resonance (DSR) technique enhances the histopathological images of cancer. Further, the DSR based Otsus thresholding processed image helps in the better segmentation of histopathological images of four types of cancer cells, i.e. breast, cervix, ovarian and prostate cancer. The comparison of segmentation results were performed on the University of California, Santabarbara (UCSB) available breast cancer datasets for analysis. The algorithm has been applied to total twenty-two (22) breast cancer images including benign and malignant and compared with Region of Interest (ROI) segmented Ground Truth images to validate the performance of proposed DSR based Otsus thresholding. DSR based Otsus segmentation obtained better results with 0.776 average correlation, 0.979 average normalized probabilistic rand (NPR) index, 0.011 average global consistency error (GCE), and 0.185 average variation of information (VI). These indices are higher than the other conventional segmentation methods and have the advantage to identify the target objects in low contrast images.
Keywords: dynamic stochastic resonance; image enhancement; tissue; segmentation; histopathological image.
A Decision Support Tool Development: An Analysis of the Statistical Significance of the Dichotic Listening of Speech Test Results
by Elena A. Popova, Evgeny L. Wasserman, Nikolay K. Kartashev
Abstract: In this study, we analyze the statistical significance of the dichotic listening test results using two methods. The first method is based on the Agresti-Coull interval, and the other one is based on the sequential Wald analysis. We also propose a way to estimate the guaranteed boundaries for the eventual laterality index at every step of the dichotic listening test using the random walk method. This work is based on the analysis of 87 dichotic listening protocols of 59 children and adolescents of 416 years of age (patients of children's psychoneurological clinic) and three healthy adult volunteers. The dichotic listening was carried out in the clinical and laboratory conditions.
Keywords: dichotic listening; speech; laterality index; random walk; lateralization; sequential Wald analysis; Agresti-Coull interval; decision support tool; statistical significance.
A Semantic Interoperability Framework for Distributed Electronic Health Record Based on Fuzzy Ontology
by Mohammed Elmogy, Ebtsam Adel, Shaker El-Sappagh, Sherif Barakat
Abstract: To achieve an efficient healthcare process; the professionals and doctors need to access the complete data about their patients in the suitable time. In the face of that issue; the medical data by natural is distributed across different systems and heterogeneous sources. From another angle; Healthcare semantic interoperability is still a huge problem without solving. In this paper, a unified semantic interoperability framework for distributed EHR based on fuzzy ontology is proposed. It consists of three main layers. The lowest layer (local ontologies construction) stores the EHRs heterogeneous data with different database schemas, standards, terminologies, purposes, locations, and formats. In the middle layer (global ontology construction) the local ontologies are mapped (using mapping algorithms or human experts with the help of common terminology vocabularies) to a global crisp one. The global reference ontology combines and integrates all local ontologies and therefore describes all data. Finally, the third layer is the user interface, in which a doctor can ask any linguistic or semantic queries by dealing with only the global reference fuzzy ontology. We expect that our framework will handle the current EHR semantic interoperability challenges, reduce the cost of the integration process, and get a higher acceptance and accuracy rate than previous studies.
Keywords: Semantic Interoperability; Fuzzy Ontology; Unified data model; DB2OWL; X2OWL.
Changes in scale-invariance property of electrocardiogram as a predictor of hypertension
by HELEN MARY M C, Dilbag Singh, K.K. Deepak
Abstract: In this study, electrocardiogram signal has been investigated to assess the presence of scale-invariance changes to classify normotensive and hypertensive subject. ECG signal of 20 normotensive and 20 hypertensive subjects is recorded using MP100 system with a sampling rate (f_s) of 500 Hz. The scale-invariance changes of ECG signal are analyzed using multifractal detrended fluctuation analysis. The width and shape of the multifractal spectrum obtained is used to detect the hypertensive subject. The multifractal spectrum of the normotensive subject exhibit a pure Gaussian behavior, but for hypertension subject is left truncated. The width of the multifractal spectrum for hypertension subject (1.7158
Keywords: Scale-invariance; Electrocardiogram; Multifractal spectrum; Hypertension.
Modified Model for Cancer Treatment
by Mahdi Rezapour Shafigh
Abstract: This paper proposes an optimal method for eradicating cancer, such that it cannot be relapsed. The major issue is that from the dynamical point of view, the tumor free equilibrium point at the end of chemotherapy is still unstable. Mathematically it means that when the chemotherapy is stoped, the dynamic behavior of the system moves away from the tumor free equilibrium point and the tumor cells starts increasing. To overcome this problem, we can either restart the process of chemotherapy or we try to stabilize the equilibrium point. In this paper the stabilization of the equilibrium point is proposed by applying a combination of vaccine therapy and chemotherapy method. According to this method, the vaccine therapy changes the dynamics of the system around the tumor free equilibrium point, and the chemotherapy pushes the system to the domain of attraction of the desired point. A suboptimal control strategy based on the State Dependent Riccati Equation (SDRE) is used for the chemotherapy process in this nonlinear cancer model. According to this method it is possible to study the particular conditions of an individual patient by choosing proper weighting matrices in the cost functions and limiting the dose of chemotherapy drug. It is shown, that, according to the simulation results, after completing the chemotherapy process, the dynamic of the system becomes stable and the cancerous cells converges to zero.
Keywords: Cancer; Chemotherapy; Lyapunov stability; Mathematical model; SDRE control; Vaccine therapy.
ASSESSMENT OF DRUG DRUG INTERACTIONS AND ITS ASSOCIATED FACTORS IN PATIENTS RECEIVING SECOND-LINE ANTITUBERCULAR DRUGS (MDR TB/XDR TB) IN A TERTIARY CARE HOSPITAL IN EASTERN INDIA
by P.Ansuman Abhisek, Priti Das, Shweta Supriya Pradhan, Geetanjali Panda, Srikanta Mohanty
Abstract: This was a prospective, observational study conducted to assess prevalence of different spectrum of DDIs in MR/XDR TB patients by analyzing 42 prescriptions. DDIs were analysed by softwares like Lexicomp, Micromedex, and Drugs.com, Medscape.com. Only interactions of major/moderate severity were included in the potential DDIs analysis. Quantitative assessments were compared using one way ANOVA. Qualitative assessment of software was done manually by mere observation. Patients received 12.52 (
Keywords: prevalence; frequency; severity; risk; pharmacodynamic; Micromedex; Lexicomp; Drugs.com; Medscape.
Peak alpha Neurofeedback training on Cognitive performance in elderly subjects
by Sofia Bobby
Abstract: Slowing down of thought, memory and thinking is a normal part of aging. Neurofeedback training (NFT) is a relatively new biofeedback technique that focuses on helping a person train themselves to directly affect brain function. In this study, EEG signal was acquired using single channel electrode, amplified using EEG amplifier and connected to the system through data acquisition device (DAQ). The peak alpha band (1011 Hz) signal was extracted using LabVIEW software. The NFT protocol that was designed presented the neurofeedback training and thereby showed some improvement in their cognitive processing speed. The visual cues were fed to the LabVIEW software by making the subject visualise some animation or hear some audios. Twenty subjects aged between 60 and 65 were considered for this training. This study had investigated whether the training given to the elderly people showed improvement in the cognitive processing speed of their brain activity.
Keywords: electroencephalography; EEG; peak alpha; neurofeedback training; LabVIEW.
Predicting Oral Squamous Cell Carcinoma in Tobacco Users by utilizing Fuzzy based Decision Tree Algorithm
by Vasantha Kavitha, Hanumanthappa M
Abstract: Oral squamous cell carcinoma (OSCC) is one of the major type of malignant and its significant percentage is responsible for the common causes of death worldwide. Many people in India have the habit of smoking tobacco and consumption of alcohol which finally leads to OSCC. Diagnosis, prediction and control of the OSCC are traditionally based on the clinical signs, historical highlights and biomarkers. In dental hospital, patient may feel inconvenient for the diagnosis on the whole ordinary methods like physical exams. The patients need differential diagnosis biopsy to predict the Oral squamous cell carcinoma. In our research, our primary goal is to predict the OSCC from the efficient decision making methods to predict the cancer from the hybrid algorithm; fuzzy based decision tree algorithm. The entire process is experimented in Hadoop framework with the mapreduce programming model. The proposed system achieves 90% accuracy in the predictions of the oral cancer.
Keywords: OSCC; decision tree; fuzzy logic; hadoop; mapreduce programming; classification.
A Frugal and Innovative Telemedicine Approach for Rural India- Automated Doctor Machine
by Aswath G.I., Shriram KV, Nalini Sampath
Abstract: Rural India is very poor [Disconnected] when compared to urban. Most of the developments have not yet reached rural India and most importantly, in Healthcare, it lacks good hospitals and sometimes they even don't have a good dispensary. Even when there are hospitals, rural people never get experienced Doctors. We know that rural people are more affected by many challenging diseases every year. To compensate all these problems and to provide a good health care even in rural villages we introduce an Automated Doctor Machine. This machine can be installed easily anywhere and also include all the required input sensors inbuilt to check the condition of the patient and also provide a diagnosis report with medicines. The patient's health report is updated to the health record in his cloud account once he receives the medicine. So, when he meets the doctor in future, the doctor can easily get his up to date health record and can also use the web app to update his health record. This is all frugal, cost-effective and fault tolerant.
Keywords: Healthcare for Rural India; Telemetry Healthcare; IOT Medical kit; Automated Doctor Machine; Affordable Healthcare;.
LUNG CANCER CLASSIFICATION USING FEED FORWARD BACK PROPAGATION NEURAL NETWORK FOR CT IMAGES
by Pankaj Nanglia
Abstract: Manual computation of lung cancer is a time taking process. In the medical industry, software aided detection (SAD) aims to optimise the classification process. This paper proposes lung cancer detection for computed tomography (CT) images. It uses speed up robust feature (SURF) for feature extraction, genetic algorithm (GA) for feature optimisation and feed forward back propagation (FFBP), neural network (NN) for classification. The training mechanism utilises 200 cancerous images and the proposed method results in 96% classification accuracy and 94.7% sensitivity. This paper also discusses the possible future modifications in the presented work.
Keywords: software aided detection; SAD; speed up robust feature; SURF; genetic algorithm; GA; FBPNN.
Study of murmurs and their impact on the heart variability
by Mokkedem Fatima, Meziani Fadia, Debbal Sidi Mohammed El Amine
Abstract: The phonocardiogram (PCG) signal processing approach seems to be very revealing for the diagnosis of pathologies affecting the activity of the heart. The aim of this paper is the application of an algorithm to explore heart sounds with simplicity in order to provide statistical parameters to better understand the cardiac activity by the calculation of the cardiac frequency and examine the impact of minor and pronounced murmurs on the cardiac variability. This paper conduct to clarify the variation of the cardiac frequency affected by several groups of murmurs and investigate the medical reasons behind the obtained results. A high number of cycles was used to better refine the expected results.
Keywords: cardiac variability; cardiac frequency; phonocardiogram signal; murmur; click; detection; duration.
Dietary Macro-nutrients Intake And Risk Of Obesity & Type II Diabetes: To compute a model to predict probability of developing Hypertrophic Obesity and Type II Diabetes based on the macro-nutrient intake levels.
by Shalini Pattabiraman, Riddhi Vyas, Shankar Srinivasan
Abstract: Good nutrition and healthy diet are the primary pre-requisites for a healthy and balanced lifestyle. Improper proportions of dietary macronutrients in the food lead to implicated risk of chronic diseases. Energy consumption and energy expenditure should be balanced in order to maintain the body mass and to avoid the complication resulting from excess body weight. Obesity is the global cause of many chronic diseases, which is why diet has an important role towards the risk of incidence of chronic diseases such as diabetes, high blood pressure and cardiovascular diseases. The purpose of this study is to compute a model to predict the risk for obesity and diabetes based on the macro-nutrient intake levels (High/Low). The SAS (Statistical Analysis Software) datasets from NHANES (2013-14) were extracted, then using descriptive statistical analysis (chi-sq & Pearsons correlation) and logistic regression analysis in SAS, a model is computed for each disease with significant predictive macro-nutrient intake variables. An ROC (Receiver Operating Characteristic) curve and predicted probability plots were analyzed for the accuracy of the model. It was found that the significant predictor variables identified for each disease risk corroborates with the literature studies, but further studies may be required with normalization of the data in order to validate the results of this study.
Keywords: Obesity; Type II Diabetes; dietary intake; macro-nutrients; prediction model; NHANES; National Health and Nutrition Examination Survey; SAS; Statistical Analysis Software; ROC; Receiver Operating Characteristics; descriptive statistical analysis; predictor variables.
A Framework for Disease Diagnosis Based on Fuzzy Semantic Ontology Approach
by Nora Shoaip, Shaker El-Sappagh, Sherif Barakat, Mohammed Elmogy
Abstract: In this paper, we propose an OWL2 ontology based on SNOMED CT standard medical terminology. It can provide significant help to support physicians in diabetes risk level diagnosis problem. It explicitly defines the semantics of diabetes knowledge by using ontology and deal with the imprecise and vague nature of its data by using fuzzy set theory. It involves building a complete linguistic fuzzy rule-base that can integrate knowledge from expert and CPG with knowledge extracted from training data and knowledge extracted from the semantic model. This step enhances the level of automation and interoperability of CDSS. The importance of our work comes from the current lack of studies related to the integration of the formal integration between the ontology semantics and FES reasoning, especially in the medical domain. The ontology acts as an integral and complementary component of the FES.
Keywords: Diabetes mellitus; clinical decision support system; fuzzy rule-based systems; semantic similarity; ontology reasoning.
A novel method of Cardiac Arrhythmia detection in ECG (Electrocardiogram) signal
by VARUN GUPTA
Abstract: ECG (Electrocardiogram) is an essential approach for observing the right condition of the heart. Generally, it represents P, Q, R, S, T and U waves. On the basis of these waves doctors can accurately diagnose cardiac arrhythmias. ECG signal is non-linear in nature and due to this, its analysis becomes very much complex and needs extra attention towards its analysis.A more reliable and accurate technique is needed for saving the life of the heart patients. So chaos theory applied to different ECG databases. R-peak is very crucial for classifying cardiac arrhythmia. For detecting R-peaks, STFT have been used and their frequency contents. In this paper physionet database has been tested.
Keywords: ECG (Electrocardiogram); cardiac arrhythmia; non-linear; Chaos theory; R-peaks.
Non-Spectral Features based Robust Speaker Independent Emotion Recognition from Speech Signal
by Revathi Dhanabal, Lakshmi Chelliah, Thenmozhi Karruppusami
Abstract: To ensure the better and effective human-machine interaction, affective computing is playing a vital role in the current scenario Since speech and glottal signals convey the information about the emotional state of the speaker in addition to the linguistic information, it is very important to recognize speakers emotions and respond to it in expressive manner This paper mainly discusses the effectiveness of non-spectral features and modeling techniques to develop the robust multi speaker independent speakers emotion/stress recognition system Since, EMO-DB Berlin database and SAVEE emotional audio-visual database used in this work contain only limited set of speech utterances uttered by ten/four actors/speakers in different emotions, it has become a challenging task to improve the performance of the stress recognition system This algorithm provides 81%and 100% as weighted accuracy recall for the stress recognition system Weighted accuracy recall is found to be 100% for the classification done on emotion specific group.
Keywords: Emotion recognition system (ERS); Vector quantization (VQ); Energy; Loudness; Zero crossing rate; Fundamental frequency; Fuzzy C means clustering (FCM); Minimum distance classifier.
A Comparative Study for Improved Hospital based Cancer Registry for Early Stage Prediction of Breast Cancer with Highest Accuracy
by Smita Jhajharia, Seema Verma, Rajesh Kumar
Abstract: Breast cancer prediction has always been a challenge and machine-learning algorithms provide great assistance in this regard. This research paper precisely reports efforts in identification and development of appropriate algorithms that can predict breast cancer with high accuracy. The research undertaken is based on rigorous analysis of data collected from Bikaner, Rajasthan, India and its quality and features in comparison to the standard SEER dataset. The results confirm the applicability of classification algorithms like Support Vector Machine in building a machine-learning model for accurate prediction of early stage breast cancer. Finally, as a highlighting contribution, a prediction model, which resulted in prediction of cancer with 99% accuracy on the data collected from Patients in Bikaner, Rajasthan, India has been presented. This model will help to improve the National Cancer Registry Program and Hospital based cancer registry systems.
Keywords: National Breast Cancer Registry; SEER Data; Breast Cancer analysis using R.
The efficacy of mechanical cervical traction for cervical spondylosis patients
by Hemlata Shakya, Shiru Sharma
Abstract: Abstract: The aim of this study is to analyze the efficacy of cervical traction for spondylosis patients using wireless EMG sensor. Cervical spondylosis is a public health issue due to dizziness, headache, and neck pain. Based on the complaint and suggest the doctors regularly for cervical traction treatment in the therapy unit. This study includes six cervical spondylosis patients for recording the EMG data using wireless EMG sensor in a sitting position. The subjects treated with 15 minutes of cervical traction with a 7 kg weight. The extracted various features in the time domain and frequency domain from the acquired EMG data to analysis the muscle fatigue during traction treatment. The pre-test evaluation showed that there are no significant differences (P>0.05) and the post-test assessment showed a very high significance (P<0.05) for outcome measurements. This statistical analysis showed that MAV, MF and MDF feature significant for spondylosis patients.
Keywords: cervical spondylosis; cervical traction; feature extraction.
Classification and Comparison of Malignancy Detection of Cervical Cells Based on Nucleus and Textural Features in Microscopic Images of Uterine Cervix
by Shanthi P.B, Shivani Modi, Hareesha K.S, Sampath Kumar
Abstract: Cervical cancer is one of the major cause of death among women worldwide. Pap smear is the cytology based screening test which is used to detect abnormal cervical cells including pre-cancerous lesions. Accurate classification of Pap smear images is one of the challenging task in medical image processing and its performance can be enhanced by extracting and selecting the well-defined features and classifiers. The irregular chromatin structure is one of the prominent diagnostic feature of abnormal cells. Classification is performed based on the extracted textural features and the benchmark Herlev dataset features. RBF (Radial Basis Function), Polynomial and Sigmoid SVM kernels are used for classification and comparison is performed with the features in benchmark database. Precision, Recall and Accuracy were calculated for all the combinations of the features. The classifier gives promising results when benchmark features are combined with textural features and also benchmark nucleus features with textural features. An effective integration of features for cervical cell classification had given good results for fast abnormal cell detection and primary Pap smear cell image classification.
Keywords: Pap smear; Cervical Cancer; Textural Feature; GLCM; SVM Kernel.
Predicting anxiety disorders and suicide tendency using machine learning: a review
by Theodore Kotsilieris, Emmanuel Pintelas, Ioannis Livieris, Panagiotis Pintelas
Abstract: Anxiety disorders constitute the largest group and the most common type of mental disorders. At the same time, machine learning techniques can be used for analyzing a patient's history and diagnose problems imitating the human reasoning or in making logical decisions. This work reviews the main concepts and applications of machine learning techniques in predicting anxiety disorder types. Seventeen (17) studies were considered, that applied machine learning techniques for predicting anxiety disorders and five (5) additional studies were examined for predicting suicide tendencies. The accuracy of the results varies according to the type of anxiety disorder and the type of methods utilized for predicting the disorder.
Keywords: Machine learning; generalized anxiety disorder; panic disorder; agoraphobia; social anxiety disorder; posttraumatic stress disorder; suicide tendency.
A novel feature set for Bone fixator classification from post-operative x-ray images
by Mrityunjaya V. Latte, Kumar Swamy V, Basavaraj S. Anami
Abstract: The paper presents a novel feature set for classification of x-ray images of bone fixators using artificial neural network. The images are obtained from radiologists. We have considered six types of bone fixators, namely, standard, ring, k-wire, screw, rod and plate. Use of both local and global features, wherein local features are defined in consultation with orthopaedicians. The feature set is reduced based on the classification accuracies of individual features. It is observed that the average accuracies for local, global and their combination are 76.83%, 66.16%and 98.3% respectively. The work finds its application in orthopaedics surgeries assisted by robots and second opinion for surgeons.
Keywords: image classification; artificial neural network; bone fracture fixators.
Automated EEG Based Epilepsy Detection Using BA_SVM Classifiers
by Aya Naser, Manal Tantawi, Howida Shedeed, Mohamed Tolba
Abstract: Epilepsy is a neurological disorder which affects individuals all around the world. The presence of Epilepsy is recognized by seizures attacks. EEG signals can provide useful information about epileptic seizures. Unlike most of the existing studies which consider only two classes, this paper proposes an automatic EEG based method for epilepsy detection which has the ability to distinguish between the three classes; normal, interictal (out of seizure time) and ictal (during seizure). In the proposed method, R`enyi entropy, line length and energy are computed from each of the five sub-bands extracted from an EEG segment using Digital Wavelet Transform (DWT). Thereafter, the extracted features are fed into BA-SVM classifiers trained using divide and conquer strategy for classification. The BA-SVM classifier is a Support Vector Machine (SVM) classifier whose parameters are optimized using BAT algorithm. The popular Andrzejak database was utilized for training and testing purposes. The average accuracies for all considered cases are more than or equal 95%. Thus, the various experiments and comparisons accomplished in this study reveal the efficacy of the proposed method.
Keywords: Electroencephalogram (EEG); Epilepsy; Digital Wavelet Transform (DWT); Entropies; Support Vector Machine (SVM); Bat Optimization.
Wavelet-Based Feature Extraction Technique for Classification of Different Shoulder Girdle Motions for High-Level Upper Limb Amputees
by Ghaith K. Sharba, Mousa K. Wali, Ali H. Al-Timemy
Abstract: There are many amputees around the world who have lost an upper-limb through accident, disease or conflict. The aim of this study is to suggest a system for classification of 7 classes of shoulder girdle motions for high-level upper limb amputees using pattern recognition (PR) system. In the suggested system, the Wavelet transform was utilized for feature extraction and Extreme Learning Machine (ELM) and Linear Discriminant Analysis (LDA) were used as classifiers. The data were recorded from 10 subjects, 6 intact-limbed, and 4 upper-limb amputees, with 8 channels involving 5 Electromyography (EMG) channels and 3- axis accelerometer. The study shows that the suggested pattern recognition system has the ability to classify the shoulder girdle motions for high-level upper limb motions with 92.67% classification accuracy for intact-limbed subjects and 87.67% classification accuracy for amputees by combining EMG and accelerometer channels. The outcomes of this study show that non-invasive pattern recognition system can help to provide control signals to drive a prosthetic arm for high-level upper limb amputees.
Keywords: Accelerometer; Extreme learning machine; Pattern Recognition; Surface Electromyography; Upper limb amputation.
Non invasive assessment of fractional flow reserve using computational fluid dynamics modeling from coronary angiography images
by S.P. Angeline Kirubha, Udaya Kumar A, Raghavi R, Reshma R
Abstract: Fractional flow reserve (FFR invasive) is measured by measuring pressure differences across a coronary artery stenosis by coronary catheterization technique. This is a gold standard method of determining the extent of stenosis. This method has some potential complications such as coronary vessel dissection, embolism, and renal failure. This paper presents a method for an assessment of FFR noninvasively, with coronary CT angiography imaging and fluid dynamics modeling. FFR is calculated as the ratio between pressure distal to stenosis and pressure proximal to stenosis of the coronary artery region segmented from CT Angiography image using MIMICS software. ANSYS software is used to determine the FFR (0.73
Keywords: Coronary CT angiography; Computational fluid dynamics; Fractional flow reserve; MIMICS; ANSYS.
An efficient ALO Based Ensemble Classification Algorithm for Medical Big Data Processing
by Saravana Kumar Ramachandran, Manikandan Parasuraman
Abstract: In this paper, we indented to propose a consolidated feature selection and ensemble-based classification strategy to diminish the medical big data. Here, the proposed system will be the joint execution of both the Ant Lion Optimizer (ALO) and ensemble classifier. So as to limit the impact of an imbalanced healthcare dataset, ALO is used for the optimal feature selection process. The optimized feature sets are classified by utilizing the ensemble classification technique. The ensemble classification method uses the diversity of individual classification models to create better classification results. In this paper, the proposed ensemble classification algorithm used the Support Vector Machine (SVM), and Recurrent Neural Network (RNN) classifier and the results of every classifier were consolidated by the majority voting technique. It was watched that the proposed ensemble technique got promising classification accuracy contrasted and other ensemble strategies. This ensemble system can administer datasets, as quick as required giving the imperative help to viably perceive the underrepresented class. The proposed approach will diminish the big medical data precisely and productively. The simulation result shows that the proposed method has better classification when compared with the single classifiers such as Random Forest (RF), SVM and Na
Keywords: Medical big data; Ant Lion Optimizer (ALO); Ensemble classifier; Support Vector Machine (SVM); Recurrent Neural Network (RNN).
Analysis and monitoring of a sensor-pill using DQPSK, with Advanced Virtual Instrumentation.
by Ajay Sharma, Hanuman Prasad Shukla
Abstract: Recording vital parameters of the internal parts of our human body is complex and challenging, so to get the exact or the accurate value we do the approximation of the data. Now to bridge this gap of accurate and approximate, technology has now taken a huge jump, and have come forward with something known as swallowable sensor pill. Although, we cant imagine of swallowing anything other than food items but still it is working. The e-pill discussed in this model already exists, and works on a complete wireless independent system, on a two-way synchronous DQPSK technique using low carrier frequency of 150-300 kHz suited for human body. This small system is incorporated in a small swallow able pill, powered by a special 3v button cell. The received body vital signals i.e. temperature, SpO2 level, pH value etc. will be live monitored and analyzed using the Biomedical toolkit of LabVIEW 8.5, in addition to it a comparative chart of more than one patient can be made for comparative analysis.
Keywords: swallowable; synchronous; DQPSK; incorporated; monitored; analyzed; LabVIEW.
Identification of Region of Interest for Assessment of Knee Osteoarthritis in Radiographic Images
by Shivanand Gornale, Pooja Patravali, Prakash Hiremath
Abstract: Osteoarthritis is the most common joint disorder in which smooth apparent on the closures of the bone wears away causing stiffness, swelling along with extreme pain. The assessment of Osteoarthritis in the early stage is most essential which is little difficult and inappropriate. The main objective of the paper is to identify the region of interest i.e. cartilage region for the detection of Osteoarthritis. In the work the database of 1173 Knee x-ray images are collected which are manually classified by two different medical experts as per Kellgren and Lawrence grading system. The histogram of oriented gradient method and Local binary pattern are used for computation. The computed features are classified using decision tree classifier. For the proposed method the accuracy of 97.86% and 97.61% is obtained with respect to Medical Expert-I and Medical Expert-II opinion. The results are promising and competitive which are validated by the medical experts.
Keywords: Osteoarthritis Knee X-ray;Median filter;Region of interest; Histogram of oriented gradients;Local binary Pattern;Decision tree.
Knee pathology diagnosis based on muscle activation intervals detection and the relationship between knee flexion and surface EMG
by Ahlem Benazzouz, Zine Eddine Hadj Slimane
Abstract: Muscle activation interval is an important clinical indicator for muscular disorders diagnosis. In this study, the S transform technique was proposed to detect the muscle activation onset and offset timing during gait and determine the relationship between surface electromyographic signals and knee pathology. The results obtained show that the proposed method achieved the shortest average latency (t_onset=0.015s, t_offset=0.014s) compared to the recent methods: SampEn, TKEO, and Integrated Profile. Moreover, the statistical analysis of activation intervals diversity and correlation between sEMG and knee flexion signals provide that for abnormal cases, the linear relationship is very weak and the activation intervals become more diverse.
Keywords: Knee pathology; muscle activation; onset and offset timing; S transform; surface EMG signals.
Improving the Prediction Accuracy of Low Back Pain using Machine Learning through Data Pre-Processing techniques
by G. Ganapathy, N. Sivakumaran, M. Punniyamoorthy, Tryambak Chatterjee, Monisha Ravi
Abstract: Application of machine learning algorithms in the healthcare industry has been increasing by many folds. Low back pain has caused problems to many persons all around the world. An early treatment or detection of whether a person has the symptoms pertaining to low back pain can help faster medication and treatment of the patient and help them with getting their medical condition degraded. This paper focuses on four different machine learning algorithms viz. SVM, Logistic Regression, K-NN and Na
Keywords: K-NN; Logistic Regression; low back pain; SVM; prediction; Naïve Bayes.
BONE METASTATIC TUMOR MINIMIZATION DUE TO THERMAL CEMENTOPLASTY EFFECT, CLINICAL AND COMPUTATIONAL METHODOLOGIES
by V.C.C. Oliveira, Elza M. M. Fonseca, J. Belinha, C.C. Rua, P.A.G. Piloto, R.M. Natal Jorge
Abstract: Bone tumors could be benign or malignant, and primary or metastatic due to systemic cancer cells dissemination. They destroy bone and lead to pathologic fractures. The main objective of this work is to study the thermal effect induced by the bone cement polymerization, in the bone metastatic tumor minimization and to understand the role of such procedure in bone tumor necrosis due to thermal effects and its biomechanical stabilization for this function. To assess the clinical effect, it is important to simulate and test this methodology before its application and obtain sustained results. In this work, a numerical model was developed to predict the temperature distribution produced by cement polymerization, and due to the thermal necrosis effect the metastatic tumor area minimization is verified. Different numerical simulations were produced for different cement sizes introduced in a cortical and spongy bone tumor, with or without an intramedullary nail in titanium. The numerical models were built according to average dimensions of patients obtained from digital conventional radiographs. For each model, the temperature distribution due to the cement polymerizing effect is represented, which affects thermal necrosis and the amount of bone cells penetration. The finite element results obtained from numerical simulations allow to conclude about the high temperature spread effect in bone material. In conclusion, values greater than 45
Keywords: Temperature; Bone tumor; Metastases; Cement; Thermal necrosis.
Real-time estimation of hospital discharge using fuzzy radial basis function network and electronic health record data
by Ahmed Belderrar, Abdeldjebar Hazzab
Abstract: Hospital resources are scarce and should be properly distributed and justified. Information about how long patients stays in critical intensive care units can provide significant benefits to hospital management resources and optimal admission planning. In this paper, we propose an approach for estimating intensive care unit length of stay using fuzzy radial basis function neural network model. The predictive performance of the model is compared to others using data collected over 13,587 admissions and 54 predictive factors from five critical units with discharges between 2001 and 2012. The proposed model compared to others demonstrated higher accuracy and better estimations. The three most influential factors in predicting length of stay at the early stage of pre-admission were demographic characteristics, admission type, and the first location within the hospital prior to critical unit admission. We have found about 63% of patients with multiple chronic conditions, stayed significantly longer in hospital. Enabling the proposed prediction model in clinical decision support system may serve as reference tools for communicating with patients and hospital managers.
Keywords: data mining; hospital administration; length of stay; machine learning; prediction model.
A NOVEL METHOD FOR AUTOMATIC IDENTIFICATION OF FOVEA LOCATION AND ITS CENTER IN COLOR RETINAL FUNDUS IMAGES
by Bharati M. Reshmi, Rajesh I S, Bharathi Malakreddy A
Abstract: The identification of fovea region plays a significant role in the detection of Diabetic Maculopathy. The presence of exudates within 500micrometer from fovea center requires laser treatment as it is a sign of sight-threatening. An ophthalmologist can decide over the nature of the treatment depending upon the position of the exudates from the fovea. Hence it is very important to identify fovea and its center with high accuracy. In this work, we have proposed a novel algorithm for identification of fovea location and its center in color retinal fundus images. The novelty of this work aims at an approach where fovea location and its center identification are done excluding blood vessels and OD. Initially, the grid is drawn over the retinal image, then the designed algorithm searches for Region of Interest (ROI) for the candidate fovea region by considering the four coordinate points. Finally, by using dark intensity property of fovea and morphological operations, fovea location and its center is identified. The proposed method is simple, robust and it is tested on a publicly available MESSIDOR database and an accuracy of 97.37% is achieved.
Keywords: Diabetic Retinopathy (DR); Diabetic Maculopathy (DM); Optic Disc (OD); Region of Interest (ROI); Blood Vessels (BVs).
Optic Disc and Optic Cup Segmentation in Retinal Images
by Thamer Al Sariera, Lalitha Rangarajan
Abstract: Automated segmentation of the optic disc (OD) and optic cup (OC) is important for retinal image analysis and retinal diabetic retinopathy systems. For OD segmentation, this paper presents a method done in three steps that combines variance and brightness features of the OD to localize leading to increased accuracy in detecting OD rather than using just one feature. As a first step, the image is divided into non-over lapping windows. Then the brightest window with maximum variance in intensity is selected. Subsequently the Circular Hough Transform (CHT) is applied to get the OD segmentation. OC segmentation,is done in two steps:(i) blood vessels inside the OD are eliminated and (ii) restricted region growing performed to get the OC segmentation.The efficacy of the proposed method is demonstrated using the standard benchmark DRIVE and DIARETDB1 databases and by comparing the results of proposed method and some methods in silico.
Keywords: Optic disc; Optic cup; Retinal image;Diabetic retinopathy; Circular Hough Transform.
A Low-complexity Volumetric Model with Dynamic Inter-connections to Represent Human Liver in Surgical Simulators
by Sepide Farhang, Amir Hossein Foruzan
Abstract: We propose a method for visualization of the human liver to represent nonlinear behavior of the tissue and to preserve the objects volume. Our multi-scale model uses dynamic interconnections to keep the size of the gland. We introduce two new parameters to control the influence of an external force on the non-linear material of the liver. Another novelty in the proposed method is to design a multi-dimension data structure which makes it possible to run our code on conventional CPUs and in real-time. We evaluated the proposed algorithm both quantitatively and qualitatively by synthetic and clinical data. Our model preserved 98.4% and 94.1% of a typical volume in small and large deformation, respectively. The run-time of our model was 0.115 second. Our model preserves the volume of a liver during both small and large deformations, and our results are comparable with recent methods.
Keywords: Medical virtual reality; Mass-spring model; Liver surgical simulator; Volumetric mesh; Multi-scale mesh model.
MR-Brain Image Classification System based on SWT-LBP and Ensemble of SVMs
by Mohammed Khalil, Habib Ayad, Abdellah Adib
Abstract: In this paper, we present an efficient MR (Magnetic Resonance) image
classification system. At the first stage, the brain image is decomposed into
several subbands using Stationary Wavelet Transform (SWT). Then, Local Binary
Patterns (LBP) with reduced histograms are computed for each subband to form
several primary feature vectors. Principal Components Analysis (PCA) followed
by Linear Discriminant Analysis (LDA) are then applied to each primary feature
vector in order to transform them into new lower-dimension feature vectors. The
third stage consists of using an ensemble of Support Vector Machines (SVMs)
in order to build voters and make the final decision on the requested image. The
designed system is evaluated on 255 brain images with Five-fold cross-validation.
Experimental results show that the proposed system achieves a classification rate
of 99.78% which outperforms the existing brain classification approaches.
Keywords: MR-Brain image classification; LBP; SWT; PCA; LDA; Ensemble of SVMs.
A systematic review on Detection and Estimation algorithms of EEG signal for Epilepsy
by Shazia Hasan, Ameya K.Kulakarni, Sebamayee Parija, P.K. Dash
Abstract: Epilepsy is the most common neurological disorder characterised by a sudden and recurrent neuronal firing in the brain. As EEG records the electrical activity of the brain so it helps to detect epilepsy of the subject. Early detection of epileptic seizure using EEG signal is most useful in several diagnoses. So aim of the work is to study and compare the different techniques used for feature extraction and classification algorithm. Epilepsy detection research is oriented to develop non-invasive and precise methods to allow accurate and quick diagnose. In this paper, we present a review of significant researches where we can find most suitable method among existing members to improve computing efficiency and detect epilepsy of the subject efficiently and accurately with lesser computational time. The database which is publicly available at Bonn University is taken.
Keywords: EEG signal; Epilepsy; Seizure detection algorithm; performance analysis; wavelet ; Hilbert transform; EMD.
Electro-pneumatic system for intussusception reduction in children and its application in the pediatric surgery
by Mohamed Ras Lain, Chouaib Daoudi, Mohamed Souilah, Abdelhafid Chaabi, Hichem Choutri
Abstract: Intussusception is an important cause of acute abdomen pain in infants and children that often occurs between 3 and 12 months of age. The early diagnosis of this disease is vital factor to avoid recourse to surgery that can poses a real danger to the infant. Recently, pneumatic reduction (Air / Gas) has emerged as a safe and promising method in the treatment of intussusception. The work presented in this article focuses mainly on the implementation of embedded instrumentation based on pressure sensor to improve the manual pneumatic system, which is currently used by the doctors. Initially, a sensor of the Motorola family (MPX 7050) is used for the acquisition of intestinal pressure. The signal conditioner is based around a specialized amplifier and the processing unit is built around Microchip PIC microcontroller. A keyboard and LCD display were used to introduce and visualize the evolution of intestinal pressure, respectively. The experimental results show the feasibility of the designed prototype in the treatment of intussusceptions.
Keywords: intussusception; pediatric surgery; pneumatic reduction; embedded system; pressure sensor; microcontroller.
Ex vivo experimental and numerical study of stresses distribution in human cadaveric tibiae
by Maria Fernandes, Elza Fonseca, Renato Natal Jorge, Maria Manzanares-Céspedes
Abstract: The mechanical behaviour of human bone tissue subject to drilling has been recently reviewed due to its increased clinical interest. However, no published data exist regarding stress analysis during the drilling. In this study, an elasto-plastic dynamic FE model of bone drilling was developed using the human cadaveric tibia obtained with a handheld 3D scanner. The FE model was validated with experimental tests and different drilling conditions were simulated in order to study the stresses distribution during the drilling process. The developed FE model was in good agreement when compared with experimental tests. Results suggest that the use of lower drill speed and higher feed-rate lead to a decrease in the stress level of the treated tibial bone. The developed FE model can be used for future studies and cover not only the mechanical behaviour of human tibiae but also the thermal aspects.
Keywords: human tibia; drilling; stresses; numerical model; experimental model; feed-rate; drill speed.
The Impact of Income Level on Childhood Asthma in U.S: A Secondary Analysis Study during 2011-2012
by Jalal AlAlwan
Abstract: Despite the abundance of researches relating children and asthma, the racial/ethnic influence on asthma threat have not been fully explained. The aim was to conduct a consistent and new study on a large-scale nationally representative data, including a minority group that has been usually eliminated form racial/ethnic literature. The 2011-2012 National Survey of Children Health (NSCH) dataset was utilized. Asthma was more prevalent among African American children (22.9%) more than white American children 13.1% (p=<.0001). Analysis of the multivariate model revealed a greater risk of asthma for the black African American children comparatively to white American children (adjusted OR" 0.522 95% CI 0.459-0.595). Our findings indicated that childhood asthma was associated with racial/ethnic status, especially with children with low income level.
Keywords: Childhood Asthma; racial/ethnic influence; National Survey of Children Health; Federal Poverty Level.
Access control to the electronic health records: A case study of an Algerian health organization
by Asma Belaidi, Mohammed El Amine Abderrahim
Abstract: Accessibility to information resources in health systems is a very important aspect. This article is about the protection of medical data and focused primarily on access control in health information systems. It is therefore a question of proposing a rigorous modelling allowing to take care of all the aspects related to the secure management of electronic health record. We proposed in the first time a model to the management of the electronic health record in the context of an Algerian health organization. Based on this modeling and by using Or-BAC model, in a second time, we proposed a model of the access control to this electronic health record. The validation of this model using the MotOrBAC tool allowed us a safe passage to an implementable specification. As a result, we develop a set of simple and effective tools to support this aspect.
Keywords: Electronic health records; Access Control; Or-BAC; MotOrBAC.
A Predictive Model for Identifying Health Trends Among Mori and Pacific People Analysis from 10-years of New Zealand Public Hospital Discharges
by Shaolong Wang, Farhaan Mirza, Mirza Baig
Abstract: Our research was focused on the quality of healthcare services for Māori and Pacific Islanders. We used New Zealand (NZ) public hospital discharges data from 2005 to 2015 for our research. A prediction model has been developed to predict the trends for patients with a specific chronic disease, external injuries and operative procedures based on the previous/historic data. Initial exploration suggests that the service demand increased from 138,656 in 2005 to 163,386 in 2015. We successfully analyzed the diseases with highest incidence rate and key characteristics of this group of patients. Our research concluded with a series of key findings on the disease types including injuries, procedures, and services.
Keywords: Predictive model; hospital discharges; machine learning model; data analysis; machine learning; predictive analysis; healthcare delivery; disease prediction; operative procedures; Māori Population and Pacific Islanders.
An effective algorithm to measure the loss of consciousness degree in epileptic seizure
by BAAKEK YETTOU Nour El Houda, Debbal Sidi Mohammed
Abstract: In this work, a new algorithm is developed to measure the loss of consciousness degrees in normal, pre-ictal, and epileptic seizure cases using bi-spectral analysis. The study is carried out on the electroencephalogram EEG signals; in which 200 records are used as pre-ictal cases, and 100 records are used as epileptic cases. All these cases are compared to 100 normal cases which represent the EEG signal in relaxed and in an awake state with open eyes. The obtained results are very satisfactory and show the efficiency of the proposed algorithm. the unconsciousness degree is very low in normal cases, very high in pre-ictal cases, and varies between high to middle during epileptic seizure cases. The algorithm promising for studying the unconsciousness degree in other cases such as anesthesia and in hypnosis cases.
Keywords: EEG signal; loss of consciousness degree; normal cases; pre-ictal cases; epileptic cases; Bispectral analysis.
R-peak detection for improved analysis in health informatics
by VARUN GUPTA
Abstract: Improvement in R-peak detection of Electrocardiogram (ECG) signal is still not saturated even requires better preprocessing, feature extraction and detection stage. Proper detection of heart diseases using the proposed technique only leads to increase its applications in medical engineering for health informatics. R-peak detection is very important for detecting heart diseases, but the involvement of various types of noises makes its detection too much complex. In this work, discrete wavelet transform (DWT) is used as preprocessing tool and Hilbert transform (HT) is used as a feature extraction tool for spectral estimation (in the form of trajectory pattern). Finally, principal component analysis (PCA) is adopted for reducing feature vectors. Detection of R-peaks is accomplished on the basis of variance obtained by first principal component (PC1). For validating this research work MIT/BIH (Massachusetts Institute of Technology/Beth Israel Hospital) Arrhythmia database has been used. The proposed technique was evaluated in MATLAB environment R2015a. The detection sensitivity (SE), positive predictivity (PP), F-score (F-s) and mean squared error (MSE) are estimated for evaluating the performance of the proposed technique. The proposed technique has resulted into SE of 99.88%, PP of 99.88%, F-s of 99.88%, SNR of 7.60dB and MSE of 0.8131%.
Keywords: Electrocardiogram; medical engineering; health informatics; DWT.
A robust Photoplethysmographic imaging for contactless heart and respiratory rates measurement using a simple webcam
by Djamaleddine DJELDJLI, Fethi BEREKSI REGUIG, Choubeila MAAOUI
Abstract: Video Photoplethysmography has been a resurgence of interest of researchers from different domains of science, driven by the demand of low cost, comfortable, contactless, non-stressful, simple and portable technology for physiological parameters measurements. In this paper, we propose a robust and simple method for remotely measure heart and respiratory activities through video Photoplethysmographic signal recordings using a low-cost webcam. The Video Photoplethysmographic signal is detected from colour video recording of a human face in an ambient light environment. The image and signal processing operations steps are minimized and optimized. Three critical aspects are endorsed during the implementation. These are performances, reduced computational time and low computational complexity. Experimental heart rate, breathing rate, and heart rate variability obtained results on 20 healthy subjects show a high correlation with those obtained using an approved contact sensor. The heart rate error obtained in the proposed method is
Keywords: ambient light; facial images; low-cost webcam; physiological parameters; Photoplethysmography; video recording.
Predicting Treatment Outcome of Spinal Musculoskeletal Pain Using Artificial Neural Networks: A Pilot Study
by Ali Al-yousef, Haytham Eloqayli, Anwar Almoustafa, Mamoon Obiedat
Abstract: Musculoskeletal pain is a heterogeneous condition with multiple risk factors, primary sources that can affect treatment and rehabilitation outcome. In this paper, we developed a prediction model for therapeutic subgrouping of musculoskeletal pain using ANN. A dataset of 27 patients with neck/shoulder pain. Patients received a single injection(0.2 ml) of 0.5% lidocaine at the trigger points.ANN model were used for predicting treatment outcome based on influential pre-treatment variables as inputs. Leave One Out Cross Validation (LOOCV) method was used for validation. The strength of each predicting variable was tested using Multilayer Feed Forward Neural Network with Back Propagation(MFFNN) and LOOCV. Then, the MFFNN prediction model was developed and designed based on the selected variables. Post-treatment endpoint follow-up(4th week VAS) was selected as a good indicator of treatment outcome. Serum vitamin D and ferritin were relatively better predictors of treatment response in the current patient group. ANN obtained 85% prediction accuracy.
Keywords: spine; neurosurgery; pain; ANN; neuron; spinal cord; AI; feature selection,DSS.
An enhanced, efficient, affordable wearable elderly monitoring system with fall detection and indoor localization.
by Ch Vineeth, Anudeep Juluru, Gudimetla Kowshik, Shriram K Vasudevan
Abstract: According to the statistics of NCOA (National Council on Aging) these days rate of death in the old aged people has reached a critical state that, for every 11 seconds an older adult is being treated for a fall and for every 19 minutes a death is reported. Most of the people leave their parents alone in the home and go for their respective jobs. In case of any fall or mishap happens to the elderly who are back at home, they are left unnoticed which may be fatal or lead to incurable disease like hip injuries, hemorrhage, tachyarrhythmia which is approximate to a cardiac arrest or even lead to death. In order to reduce these types of risks faced by elderly people, we designed an affordable IoT wearable product that can monitor their movement, the health of old people which can detect their fall immediately. Falls inside the bathroom may be fatal even for young, healthy people. Most of the elderly persons tend to remove their wearables before entering the bathroom and one cant force them either. In order to detect falls even without wearable band, we designed a smart bathroom which is capable of detecting falls and alerts when a fall occurs. These type of bathrooms can be installed in houses, hotels in order to ensure their customer safety. Deaths due to unnoticed falls inside the bathroom can be prevented by installing our system, which will alert the respective authorities immediately when a fall is detected.
Keywords: Fall detection; Old age support; IoT for medicine; Old aged tracking system; Android App; Bathroom fall; Indoor Localisation.
Performance evaluation of a computational model for brain shift calculation
by Karin Correa, Natividad Bermejo, Oscar Andrés Vivas, José Maria Sabater
Abstract: This article shows a solution for the computation of deformable tissues displacements in the brain shift problem in neurosurgery. In this type of surgery, the brain moves and deforms, changing the pre-surgical reference the surgeon had before the intervention. Among the causes of brain shift are the effect of gravity, loss of cerebrospinal fluid as a consequence of the resection practiced, the effect of the drugs supplied, among others. This document refers to the physical model of this displacement to later simulate them in multiphysics software. A phantom test was constructed by means of hydrogels, imitating the porcine brain tissue, which is subjected to compression along the z axis. The results show that the simulation proposed reproduces the behavior of the real phantom with a high level of accuracy. The application developed may serve in the future to reproduce the total behavior of the brain and thus obtain a better calculation of the brain shift.
Keywords: Neurosurgery; brain shift; neuronavigation; medical robotics; constitutive models; hyperelastic material model; constitutive parameters.
A new design of real-Time monitoring and spectral analysis of EEG and ECG Signals for epileptic seizure detection
by Boumedyen BELAID, Zine-Eddine HADJ SLIMANE
Abstract: The evolution of telecommunications technology has made significant contributions and advances in medical technology. Most of the time, monitoring and evaluation require the use of more than two signals simultaneously recorded. Simultaneous monitoring of the electrocardiogram (ECG) and the electroencephalogram (EEG) is very useful to have information about general state of health of the patient. In this paper, a novel mono-channel wireless ECG&EEG system for epileptic seizure detection is presented. The system employs analog circuits to acquisition, processing and spectral analysis of ECG and EEG signals simultaneously. Arduino Platform is used to digitize and spectral analysis of signals. A 128x64 Graphic LCD Display module and a Bluetooth module are also used for plotting and transmission of signals. We also propose the magnitude squared coherence (MSC) as an important parameter to calculate in frequency domain the relationship between ECG and EEG signals and use it as a relevant discriminator in seizures and the epilepsies classification.
Keywords: Electrocardiogram (ECG); Electroencephalogram (EEG); Arduino; Bluetooth; 128x64 Graphic LCD Display Module; magnitude squared coherence (MSC); Epileptic seizure detection.
Special Issue on: Health Engineering and Informatics
Bio-Medical Analysis of Breast Cancer Risk Detection based on Deep Neural Network
by Nivaashini M, Soundariya R S
Abstract: - Breast tumor remains a most important reason of cancer fatality among women globally and the pathetic condition is that most of them pass away due to delayed diagnosis. But premature recognition and anticipation can significantly diminish the chances of death. Risk detection of breast cancer is one of the major research areas in bioinformatics. Bioinformatics can be narrowly defined as a field at the crossroads of biology and computer engineering, responsible for the storage, distribution, and analysis of biological information. Various experiments have been conceded to examine the fundamental grounds of breast tumor. Alternatively, it has already been verified that early diagnosis of tumor can give the longer survival chance to a patient. Consequently, early prediction of breast tumor is very critical to rescue a patients life. To deal with this issue, in the proposed framework a model with deep learning techniques has been introduced to identify breast tumor at primary stage. This paper aims at finding an efficient set of features for breast tumor prediction using deep learning approaches called Restricted Boltzmann Machine (RBM). The proposed framework diagnoses and analyzes breast tumor patients data with the help of Deep Neural Network (DNN) classifier using the Wisconsin dataset of UCI machine learning Repository. and, thereafter assesses their performance in terms of measures like Accuracy, Precision, Recall, F-Measure, etc.
Keywords: Breast tumor Prediction; Feature Selection; Restricted Boltzmann Machine; Deep Neural Network.
A REVIEW ON EMOTION RECOGNITION IN PARKINSONS DISEASE USING BIOINFORMATICS
by Rejith K.N., Kamalraj Subramaniam
Abstract: Parkinsons disease individuals have been stressed and shown difficulty in emotion recognition & facial expression with increasing cognitive decline. In recent years, various studies have been conducted in emotion recognition of Parkinsons disease patients. In many research works, emotional state assessment using facial expression and EEG based stimuli were used for emotion recognition study. In this paper, a review of EEG based emotion recognition in Parkinsons disease and its various research analyses in the past two decades have been analyzed. Most of the papers have investigated total of six emotions in Parkinsons disease study such as happiness, sadness, fear, anger, surprise, and disgust to evaluate the emotion difficulties in Parkinsons disease people.
Keywords: Cognitive deficit;Electroencephalogram;Emotion; Emotional deficits; Event related potencial;Facial emotion recognition; Non-linear methods; Parkinson’s disease.
QUERY OPTIMIZATION WITH WEIGHTED FISH SCHOOL SEARCH IN ONTOLOGICAL DATABASE WITH APPLICATION OF BOINFORMATICS
by Jaya Raju, C.S. Pillai, R. Jagadeesh Kannan
Abstract: Making queries from large ontological database has a severe problem of generating query plans as it is made in form of left tree search form. This restricts the querying for composite applications and speed of acquiring query results. In such a scenario the most prominent approach is to optimise the indexing of graph nodes in ontological database and many evolutionary and particle of swarm optimisation (PSO)-based approach had already been attempted. However, loss of diversity and unanticipated convergence causes the solution to remain sub-optimal. In this study we present a weighted fish school searching-based query optimisation technique owing to its scalability and self control functioning with the application of bioinformatics. It creates probabilistic logic-based weight system for the fish school search in a hierarchical tree form which results in increased accuracy when put in comparison with standard PSO-based methods and its other variants.
Keywords: optimisation; semantic data; particle of swarm optimisation; PSO.
A cloud based secured Framework for smart Medical Diagnosis: A Survey
by Leelavathy Mrs, Selva Brunda S
Abstract: Enormous amounts of medical data are being collected by several well-developed HIS (Hospital Information Systems) in the form of patient records in hospitals. The hidden patterns and relationships contained in this data are identified using various data mining techniques which has drawn increasing world-wide attention in the recent years. As a result of which, a good number of MDSS (Medical Decision Support Systems) have been developed. These are computer systems designed to assist physicians or other healthcare professionals in making clinical decisions for the given patients symptoms and medical history. Inspired by the existing systems, the model proposed in this paper aims at making a knowledge sharing collaborative platform for doctors which serves as a realistic and effective Medical Decision Support System. It is important because it provides vital information from different data sources. It has several challenges like scalability, response time , heterogeneous data formats . This paper does a detailed study of different challenges in understanding about medical information that can be provided to users in a better manner . A framework named Intelligent Health care Framework(ISHF) is being proposed in this paper.
Keywords: Medical decision support system (MDSS); Bioinformatics; diagnosis; knowledge discovery and symptoms.
A Study on Indirect Immunofluorescence Image Classification Methods for Bioinformatics
by Divya BS, Kamalraj Subramaniam, Nanjundaswamy HR
Abstract: The Indirect Immunofluorescence (IIF) test with Human Epithelial type-2 (HEp-2) cells as substrates is the gold standard for Anti Nuclear Antibodies (ANA) test to diagnose autoimmune diseases. The specialists in the laboratory visually examine the specimen under microscope to recognize the staining patterns and generate the report. So ANA test is subjective and needs systemic automation for bioinformatics. In this view international benchmarking initiatives were organized by IAPR in the last six years. In this paper the state of the art on IIF HEp-2 cells classification task was analyzed. This paper highlighted the original aspects with the detailed discussion of the published methods. Design choice verses performance was analyzed.
Keywords: Anti Nuclear Antibodies; pattern classification; HEp-2 cell,Computer Aided Diagnosis; Indirect Immunofluorescence; Healthcare System.
Special Issue on: Biomedicine in Industry and Society
ALLIUM SATIVUM - BREAST CANCER THERAPEUTIC AGENT TO REPLACE ALLOPATHIC TREATMENTS WITH EXTENSIVE SIDE - EFFECTS
by Neha Vutakuri, Sita Somara
Abstract: Cancer is a life-threatening disease that can grow in any part of the human body and may result in the abnormal growth of millions to trillions of cells. There are currently 200 types of human cancers. Cancer is treated with various treatments, such as surgery, radiation therapy, chemotherapy, immunotherapy, stem cell transplantation, targeted therapy, and others. However, these treatments produce side effects, such as anemia, diarrhea, appetite loss, bleeding, hair loss, infections, lymphedema, nerve problems, urinary, bladder problems, and others. The side effects of cancer treatments may also lead to death. Some spices are used to naturally fight cancer. Spices are usually used in cooking as a flavoring ingredient and may have thousands of medicinal characteristics. Spices such as turmeric, ginger, garlic, saffron, black pepper, cayenne pepper, mustard, mint leaves, black cumin, and others are used in Ayurveda medicines for the natural treatment of cancer. This paper addresses the use of garlic (Allium sativum) in the treatment of breast cancer and its promising results. Garlic is a medicinal plant and an antibacterial agent that provides many benefits to humans if consumed in the regular diet. Garlic has been used for thousands of years to treat several diseases and is known to inhibit breast cancer growth. Many epidemiological studies have found that garlic is one a natural remedy that inhibits the growth of breast cancer cells. The dietary consumption of garlic in various forms effectively minimizes the growth of breast cancer and other diseases. This paper provides a complete description of garlic including its history and then discusses its utility for preventing of breast cancer and its use in clinical trials; hence, this paper will be helpful for researchers and physicians in treating breast cancer.
Keywords: Breast cancer; Spices treatment; Garlic; Ayurveda; Antibacterial agent.
Investigation of Problems faced during capturing of Gait Signals
by Vinothkanna Rajendran, Prabakaran Narayanaswamy, Sivakannan Sivakannan
Abstract: To solve the issue of close human contact in biometric authentication system, relatively a new technique gait recognition is used. The human gait is a common feature for identifying the walking manner of the person during walking and it is viewed as significant indicator for gait function of individual for experimental and research setting. Gait symmetry is usually considered as function of locomotion between the changes of the body and its activities. An exclusive advantage of gait as a biometric is its latent for detection at a distance or at low resolution or when other biometrics might not be perceivable. In this paper we investigate the problem of people recognition by their gait. The coordination and cyclic nature of the body motion makes gait as unique characteristics of each individual, thus a good biometric identification approach. The purpose of this paper is to describe some of the problems that make it difficult to apply gait as a biometric identification and use recent literature to show suggestions being made to solve some of these technical issues.
Keywords: Gait recognition; biometrics; walking style; pathology; injuries.
Improving the Classifier accuracy with an Integrated approach using Medical data - A Study
by Maragatham Ganesan, Rajendran S
Abstract: As information plays a vital role in the current scenario, Fetching of information from the voluminous quantity seems to be challenging. Therefore, the data mining community work on this area to find an improved solutions to help the end users. The end users may be an organization or may be an ordinary user. The authors have used different classification techniques for the study purpose. The article attempts to analyze the accuracy of classifiers with respect to that of medical data. Dataset from the repository is considered for analyzing purpose. Initially a preprocessing step is used on the data set for finding out the missing values. Next, the resulting data set is applied to the classifiers to study its performance accuracy. Initially, after the preprocessing step a classifier is selected based on which the accuracy of the classifier is studied. Next, a subset evaluation step is performed to determine the relevant attributes. In order to improve the classifier accuracy an attribute selection filter of supervised category is selected. In this article, the authors have done a study on integrated approaches that helps in classifying the instances of a bio-medical data. For the analysis purpose the Na
Keywords: Naive Bayes; Multilayer perceptron; chi-square; logistic regression.
Wavelet Packet Transform based Medical Image Multiple Watermarking with Independent Component Analysis Extraction
by Nanmaran Rajendiran, Thirugnanam Gurunathan, Mangaiyarkarasi Palanivel
Abstract: Rapid growth of internet in all aspects of life has led to the easy availability of the digital data to everyone. E-commerce, telemedicine etc., are among the many applications of internet. Telemedicine is a crucial field where internet finds application. Healthcare professionals use internet to transmit and receive medical data. Thus the medical images can be shared, processed and transmitted through computer networks. All patient records, linked to the medical secrecy, must be confidential. Because of the importance of the security issues in the management of medical information, there is a need to develop watermarking techniques for protecting medical images. In this paper, colour medical image watermarking methods rely on Wavelet Packet Transform (WPT) and extraction using Independent Component Analysis (ICA). For watermark extraction, Pearson ICA is applied as it attains the new trait is that it not entail the renovation procedure in watermark extraction. The grades show that projected method is vigorous beside attacks such as Gaussian noise, Salt and Pepper noise, Rotation and Translation. The performance measures like PSNR, Similarity Measure and Normalized Correlation are assessed to confirm the robustness of the scheme.
Keywords: discrete wavelet transform; image watermarking; independent component analysis; pearson ICA; wavelet packet transform.rn.
Novel feature extraction of EEG signal for accurate event detection
by S. Saravanan, S. Govindarajan
Abstract: Electroencephalograph (EEG) signal analysis is one of the essential for diagnosis of diseases, detect the various event and psychological problems. The accuracy of event detection depends on the feature vectors used. The feature vectors in literature provide performance only for a specific application and perform poor for other application. This paper presents a new feature vector generation using fusion of energy feature vectors of different types. The energy features of alpha, beta and gamma component is extracted as base feature. A fusion rule is formulated to fuse that base features using Hjorth activity features to improve the accuracy of classification. The performance of the proposed new feature is tested on seizures detection, sleep state detection and emotion detection application. The resulting analysis shows that the proposed new feature outperform with improved accuracy of 26% for emotion detection, 5% for seizures detection and 4% for sleep state detection.
Keywords: electroencephalograph; EEG; event detection; energy feature; emotion detection; feature extraction; fusion; Hjorth activity; seizures detection; sleep state detection.
OPTIMIZED FEATURE SELECTION AND ENTROPY-BASED GRAPH CLASSIFICATION OF GENE EXPRESSION DATA
by Audu Musa Mabu, Rajesh Prasad, Raghav Yadav
Abstract: Purpose: Gene expression (GE) profiles expansively revised to disclose intuition into the multifariousness of cancer furthermore to discover concealed information which provides biological knowledge for the classification of cancer. Precise cancer classification straightly through original GE profiles stays challenging on account of the intrinsic high-dimension feature along with the small magnitude of the data samples. Therefore, choosing high discriminative genes as of the GE data have turn into progressively fascinating in the bioinformatics field. This given paper gives a technique for the GE data classification utilizing entropy based graph classifier.
Methodology: Initially, the proposed technique evaluate the GE datas SNR (Signal to Noise Ratio) values, additionally, selects the relevant features using KH (Krill Herd) optimization process. The truth is that not all features are helpful for classification, and some redundant together with the irrelevant features might even serve as outlier. To dispose the outliers, feature reduction is done with the assist of Euclidean distance. Classification is made utilizing entropy based graph classifier.
Conclusion: The proposed process effectiveness contrasted with the existing method concerning classifications is established from the Experimental outcome.
Keywords: Gene Expression Profile; Bioinformatics; Entropy; Graph Classifier; Krill Herd Optimization; Signal to Noise ratio.
Novel Multiphase Contouring and Force Calculation Algorithm for ROI Detection and Calculation of Energy value in Multiple Scale and Orientation for Early Detection of Stages of Breast Cancer
by Varalatchoumy M, Ravishankar M
Abstract: A novel MCFC algorithm has been developed to perform detection of ROI. Detected malignant tumors were processed using a novel approach to identify stages of breast cancer. Preprocessing phase aids in enhancement and noise removal. Preprocessed image is segmented using the MCFC algorithm to detect the ROI that aided in achieving robust segmentation at very low computation time. Combination of wavelet and textural features were used to train the artificial neural network for classification. Tumour stage is identified using a novel approach of calculating the energy values in four different scales and six orientations for each scale. Total of 24 energy values are used for training. System performance was tested on 45 real time patients mammographic images obtained from hospitals. Detection of malignant tumour and its stages was verified by experts in medical field. Overall accuracy obtained is 97% for MIAS images and 90% for real time mammographic images.
Keywords: multiphase contouring and force calculation; MCFC; stages of breast cancer; wavelet and textural features; energy values.
An Amalgamated Prediction Model for Breast Cancer Detection using Fuzzy Features
by Smita Jhajharia, Seema Verma, Rajesh Kumar
Abstract: Machine learning techniques for cancer prognosis are being widely researched and applied. Supervised learning in conjunction with other computational techniques are paving the way forward for predictive health analytics. The step of input feature processing is very important in order to obtain meaningful results from a data analytics problem. In this paper, the Extended Kalman Filter (EKF) and Fuzzy K-Means clustering algorithms have been combined and a hybrid algorithm is proposed with improved functionality compared to either of the two separately. The proposed hybrid algorithm implements a fuzzy K-means algorithm with Support Vector Machine (SVM)coupled with an EKF for data filtering. From the filtering and prediction consecutive cycles, the result of Kalman filters is obtained and fuzzy membership functions are calculated in order to form a relationship between the labels and attributes. K-means utilizes this relationship to create a new modified set of attributes, which are given to the SVM classifier, with lesser number of support vectors. The number of clusters is added into the training process as the input parameter except the kernel parameters and the SVM penalty factor. The approach was tested for various publicly available data sets like UCL,SEER and a real data set compiled by the authors. After performing statistical analysis, the accuracy, precision, recall and F-score value of the algorithm have been found and compared against those obtained from the traditional algorithms.
Keywords: Cancer; Clustering; Extended Kalman Filter; Fuzzy K-Means.
EEG signal analysis and classification on P300 speller-based BCI performance in ALS patients
by Mridu Sahu, Shrish Verma, Naresh Kumar Nagwani, Sneha Shukla
Abstract: Objective: the objective of the presented work is to analyse the electroencephalography signal based on brain computer interface by using P300 speller for amyotrophic lateral sclerosis (ALS) patients and perform classification on extracted features to get accuracy. Analysis/methods: amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that involves the degeneration and death of nerve cell in the brain. It affects the process related to speech and loss of motor function in the patient. BCI technology is a communication solution for all amyotrophic lateral sclerosis (ALS) patients. The P300 speller included in the BNCI Horizon 2020 data is an application allows calculating the accuracy of classifier, which is necessary for the user to spell letters or sentences in a BCI speller paradigm. In this paper, we have extracted wavelet and power spectral density features. Association rule mining and ranking method is used for feature selection. For the classification, we have used multiple techniques and different classifiers and out of those, ten best techniques are selected based on their good performance. Finding: as a result, we get maximum 75% accuracy when we used random committee classifier.
Keywords: amyotrophic lateral sclerosis; ALS; brain computer interface; BCI; electroencephalography; EEG; P300 speller; power spectral density; PSD; wavelet; association rule mining; ARM.