OPTIMIZED CLASSIFICATION OF COLORECTAL CANCER USING DEEP LEARNING TECHNIQUES

Abstract

The healthcare sector is quite different from other sectors. Expectations of people are of greatest level in terms of services in this area regardless of cost. A large amount of budget is being spent on these services but still the expectations of people are not met. The medical field has various modalities which produce large amount of medical images. The information hidden in an image is worth more than a thousand words. Proper analysis of a medical image can help in timely detection and diagnose of a disease which increases the rate of survival of patients. These images are read or interpreted by human experts. But interpretation by human experts has limitations because of reasons like image complexity, image subjectivity, variations among different interpreters and fatigue. So, there is need of more appropriate method for interpretations. Deep learning has given successful solutions for many real world applications. It also gives solution for medical image interpretation with high accuracy. But, penetration of deep learning in healthcare industry is quite slow as compared to the other real world problems because of the sensitivity of the domain. We have highlighted the issues, which act as barrier in the growth of deep learning in the health care industry. We have also discussed the application areas of deep learning in medical image analysis. We have also presented the overview of the work done in this field.