SYSTEMATIC REVIEW OF RECENT COMPUTER AIDED DIAGNOSTIC APPROACHES IN MEDICAL IMAGING

Computational Intelligence is impacting many areas of daily life. Health is one of them. One emerging technology of Al, machine learning is being utilized to diagnose the disease present in the patient. This area which deals with computational Intelligence techniques by analyzing scanned images of the affected body part by the suitable imaging device to detect the presence or absence of disease is known as Computer-assisted diagnosis. These scanned images are of very high resolution still it is difficult with just by visual examination for any human to conclude about the disease in the patient. Research is this direction is more than three decades but still not widely popular and acceptable among community of radiologists and physicians. A number of approaches have been suggested for Computer assisted diagnosis for different set of diseases and imaging modalities. CAD systems use trained datasets for classification. These training datasets are prepared by team of experts who manually annotates images. These CAD systems processes the images and highlights image segments where there is greater degree of suspicion of lesions which are further processed by radiologists for final conclusion. Such systems save a lot of time and effort of radiologists and also can increase their efficiency. In this paper, authors have presented systematic literature review of recent approaches by keeping focus on critical performance issues and also presented future scope of machine learning approaches which could be used in computer assisted diagnosis.