International Journal of Computational Medicine and Healthcare (9 papers in press)
A Cellular Automaton-Based Passive-Acoustic Technique for Topological Characterization of Objects in Fluid With Potential Application to Carotid Artery Plaques
by Matthew Delaney, Gillian Pearce
Abstract: We present a system for the detection and characterization of objects located in tubes of flowing fluid. Our system makes use of Semiotic Analysis of a Driven Greenberg-Hastings Cellular Automaton. Our results indicate that this system is highly effective in both detecting and distinguishing topological features of different objects. We conclude that we have created a potentially effective and relatively low cost system with strong potential for the detection and initial analysis of plaques causing occlusions in the carotid artery. We further conclude that the resulting system could potentially be used as a front-line screening method in medical settings such as GP surgeries and clinics.
Keywords: carotid; artery; plaque; cellular automaton; complex systems; semiotics.
A Systematic Literature Review of Data Forecast and Internet of Things on the E-Health Landscape
by Gabriel Souto Fischer, Rodrigo Da Rosa Righi, Vinicius Facco Rodrigues, Cristiano André Da Costa
Abstract: Internet of Things (IoT) is a constantly expanding paradigm that promises to revolutionize healthcare applications and could be associated with several other techniques. Data forecast is another paradigm widely used, where data captured over time are analyzed in order to identify and predict problematic situations that may happen in the future. After research, we did not find surveys that address IoT combined with Data Prediction in healthcare area in literature. In this context, this work presents a systematic literature review on Internet of Things applied to E-Health Landscape with a focus on Data Forecast, presenting as results fourteen papers about this theme, and a comparative analysis between them. Our main contribution for literature is a taxonomy for IoT systems with Data Forecast applied to healthcare. Finally, this paper presents some possibilities and challenges of exploration in the study field, showing existing gaps for future approaches.
Keywords: Internet of Things; Health; Data Forecast; Sensors; Distributed Systems; Taxonomy; healthcare Environments; Survey.
Simulation and analysis of Hiv-AiDs dynamics
by Jimbo Henri Claver
Abstract: One of the most intriguing questions in mathematical epidemiology is how can one efficiently control and prevent the propagation of a disease. The problem of disease modelling, simulation and control becomes even more fascinating if we look at various risk groups. Referring to Hiv-Aids disease, it is worldwide agreed that the HIV virus seemingly knows when it should attack the body such as to develop AIDS disease. The fundamental question is therefore related to the time and location of such process to happen. To answer to this question, we study a model of propagation of HIV-AIDS in a given population. The AIDS disease is hardly easier to understand than HIV propagation dynamic, but fortunately, we can simplify the system even further by studying the susceptible and infected population dynamics in their behaviour in isolation and/or interaction. Finally, we develop a simulation model based on observed behaviours of susceptible and infected populations. This allows us to test our ideas of how the HIV virus develops into the AIDS disease within the highly controlled environment of computer simulation. Based on these insights, we can suggest new experiments on the actual system and update our models accordingly.
Keywords: Simulation; disease modelling; management; parameter specification; biodynamic; noise biology; biomathematics; computational biology; Hiv-AiDs modelling; stability and computation.
Integrating the Kano model for optimizing CPR-D training system
by Jingjing Liu, Huijun Xi, Li Gui
Abstract: Purpose: To explore customers requirements of CPR-D training system and optimize the system accordingly. rnMethods: We conducted a Kano model-based questionnaire survey among medical staffs with 28 quality features of the CPR-D training system being developed earlier. A modified Kano categorization was adapted to decide the final category.rnResults: Totally 268 of the 300 questionnaires distributed were valid. Most of the participants were either physicians or nurses, while the rest were non-clinical medical staffs or nursing teachers. Of 28 features, 4 were attached to attractive attributes, 15 were one-dimensional attributes, 7 were indifferent attributes, and one was reversal attribute, while one feature was ambiguous. After the modified Kano categorization, 4 were categorized to attractive attributes, 19 were one-dimensional attributes, 4 were indifferent attributes, and one was reverse attribute. Comprehensively considering the results, version 1.0 of CPR-D training system was upgraded to version 2.0, in which a total of 7 QFs were optimized.rnConclusion: The Kano model-based questionnaire provides valuable information for optimization of CPR-D training system. In the future, continuous survey should be conducted to update customers requirements.rn
Keywords: Kano model; CPR; defibrillation; training.
A comparison of structured data query methods versus natural language processing to identify metastatic melanoma cases from electronic health records
by Jinghua He
Abstract: The relative efficacy of natural language processing (NLP) of text reports compared to structured data queries for identifying patients from electronic health records (EHRs) with metastatic cancer remains unclear. Such identification is critical for identifying and recruiting potential study candidates for cancer trials, particularly trials of cancer chemotherapy. For such purposes, we performed a direct comparison between NLP and structured data query methods for identifying patients with metastatic melanoma. Using EHR data from two large institutions, we found that NLP of text reports identified close to three times as many patients with metastatic melanoma compared to a structured data query algorithm (1,727 vs 607 patients). Using an external tumor registry, we also found NLP had much higher sensitivity than structured query for identifying such patients (67% vs 35%). Our results emphasize the importance of employing NLP criteria when identifying potential cancer study candidates with metastatic disease.
Keywords: efficacy; natural language processing; structured data queries; identification; identifying patients; electronic health records.
Extraction of Breast Cancer Biomarker Status using Natural Language Processing
by Paul Dexter
Abstract: The primary objective for this project was to determine the performance of natural language processing (NLP) algorithms for extracting estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor 2 (HER2) receptor status for patients with breast cancer using unstructured (free text) EMR data, and to determine the prevalence of triple negative breast cancer in the Indiana Network for Patient Care (INPC) population. After obtaining IRB approval, structured queries of the INPC data repository were performed to identify female patients with a history of breast cancer over a 10 year period who had at least five oncology notes or one related pathology document. Using this patient cohort with evidence of breast cancer and related clinical documents, we employed these patients' unstructured text EMR data to develop NLP algorithms to detect ER, PR, and HER2 receptor status. To support the algorithm evaluation and iteration process, we performed manual review and annotation of a random set of target documents. The performance of our NLP algorithms for extracting ER, PR, and HER2 receptor status was good with sensitivity 87.5% to 92.6%, specificity 88.6% to 95.8%, positive predictive values (PPV) 82.4% to 99.0%, and negative predictive values (NPV) 85.2% to 97.7%. This study confirmed our primary hypothesis that NLP algorithms are effective in identifying important breast cancer biomarkers in patients with breast cancer using unstructured data.
Keywords: NLP algorithms; Effective; Breast cancer biomarkers; Breast cancer.
Measuring the Impact of Certified Electronic Health Record Technology on Cost, Quality and Safety Outcomes
by Joseph G. Conte
Soup Burns and the Roles Played by Viscosity, Solid Constituents, Epidermal Thickness and Clothing
by Hunter Hickman, Jared Powell, Tony Kerzmann, Alexandra Cox, Gavin Buxton, Carl Ross
Abstract: One of the many concerns for parents of young children and caregivers of geriatric patients is the potential for scalds from hot food or beverages. Because the skin tissue in young and elderly people can be thinner and more susceptible to scalds, and soups are a common food for young and elderly people, the accidental spilling of hot soups can be a common source of burns for these patients. The type of soup, however, may not elicit any such misgivings, with parents of young children and caregivers of elderly patients not aware that different types of soups could result in more or less severe scalds. Here we elucidate the effects of reduced skin thickness, as is often found for young children and geriatric patients, on the severity of soup burns. In particular, we compare the two most popular types of soups (tomato and chicken noodle) along with hot water, to identify the roles that the viscosity and solid constituents in these different soups may play in the severity of the burn. We find that the more solid constituents in the chicken noodle soup, in particular, prevents the soup from flowing from the skin, which increases the time the skin is exposed to the elevated temperatures of the soup, and therefore increases the severity of the burn.
Keywords: Computational Dermatology; Computational medicine; Computational Burn Severity; Skin Burns; Geriatric Burns; Pediatric Burns.
Investigation of normal and abnormal blood pressure signal using hilbert transform, Z-transform and modified Z-tranform
by VARUN GUPTA
Abstract: The analysis of neurocardiological signals and their fluctuations is a popular tool of medical diagnostics. Aside from heart diseases, it may investigate the performance of many other organs. In this category, the electrocardiogram (ECG), blood pressure, respiration rate and blood flow are the most versatile signals. Blood pressure (BP) control is very significant in the lifespan of every individual. If it goes out of the boundary, then runs a risk of heart attack increases. During anesthesia, BP is the main challenge. If any time, lag time (dead time) comes in the process and then it is very unsafe for the life of that person. The first measure towards the minimization of that delay time is its analysis. Delay time may be integer or fractional in nature. That is why ordinary Z- transform fails in that category. In this paper authors have been proposed modified Z-transform and its execution on real hospital BP database. Modified Z-transform gave better result at low values of the sampling period. Transient and steady-state parameters were also checked for that analysis. Hilbert transform is used to find out the hidden time information (on the basis of obtained frequency contents).
Keywords: Blood Pressure; sampling period; modified-Z transform; Delay time.