Title: An optimised multidimensional feature vector for automatic tumour detection and retrieving similar images for diagnosing clinically
Authors: B. Jyothi; Y. Madhaveelatha; P.G.K. Mohan; V.S.K. Reddy
Addresses: Department of Electronics and Communication Engineering, Malla Reddy College of Engineering & Technology, Hyderabad, India ' Department of Electronics and Communication Engineering, Malla Reddy Engineering College Women, Hyderabad, India ' Department of Electronics and Communication Engineering, Institute of Aeronautical Engineering, Hyderabad, India ' Department of Electronics and Communication Engineering, Malla Reddy College of Engineering & Technology, Hyderabad, India
Abstract: High speed advancements in the medical technology are revolutionising the medication. The usage of Computer-Aided Diagnosis (CAD) systems plays a potential role in improving the quality of clinical radiotherapy care. The tumours are fatal in nature and may invade the other parts of the body. Precise diagnosis is essential for accurate treatment of the tumour. In that direction to make easy access in the future all medical images are registered in a medical image database. Content-Based Medical Image Retrieval (CBMIR) techniques attract greater interest to facilitate medical fields in retrieving the interested medical images. Consequently, based on this, we propose an intellectual three-step CBMIR system which retrieves corresponding images from a large medical database to support the radiologist in the diagnosis tumour images. The initial step detects the tumour in the image. Based on this information it retrieves the nearby similar images within the pre-classified database with the optimised futures. In the final step, the performance of the retrieval system is enhanced by adapting the subjective feedback method. The experimental results show that the proposed region-based multiple features and integrated with classifier and subjective feedback method yield better results than classical retrieval systems.
Keywords: content-based medical image retrieval; tumour detection; relevance feedback; feature extraction.
DOI: 10.1504/IJBET.2017.085138
International Journal of Biomedical Engineering and Technology, 2017 Vol.24 No.3, pp.207 - 224
Received: 12 Mar 2016
Accepted: 16 May 2016
Published online: 13 Jul 2017 *