A Survey of Medical Image Analysis Using Deep Learning Approaches

With the expanding development of Deep Learning techniques Medical Image Analysis have become an active field of research. Medical Image Analysis typically refers to the utilization of various kinds of image modalities and techniques to obtain images of the human body which in turn can be used by medical experts for diagnosis along with treatment of patients. This paper provides a survey of various improvements that have been made in Medical Image Analysis using DL techniques related to different pattern recognition tasks. These pattern recognition tasks include Classification, Detection/Localization, Segmentation, and Registration. . The paper discusses several recently published research papers related to different pattern recognition tasks including liver lesion classification and segmentation, lung nodule detection & classification, lung nodule segmentation, brain tumor classification, and detection, brain tumor segmentation, Breast cancer detection, etc. Comparative description of these papers is also provided in terms of organ, modality, dataset, model used and limitation/improvements needed. This survey briefly describes several medical imaging modalities used in medical image. Also, the proposed research work has evaluated various challenges encountered in the Medical Imaging domain and have discussed about the current trends for new researchers/ medical instrument experts encouraging them to take full advantage of DL techniques in the future.