Deep Learning Challenges in Medical Imaging

Abstract

Medical imaging has taken decisive steps in diagnosing the diseases in the past few decades. Many successful diagnosis systems have been developed, and however, there is a scope for improving the accuracy of those systems. Deep learning plays a vital role in improving the performance of medical imaging. Recent developments in deep learning have brought revolution in medical diagnosis. Recent progress in deep learning offers new possibilities in medical imaging for detection of diseases, diagnosis and prediction. Use of deep learning to medical imaging has remarkable achievement in improving the performance in diagnosis. Deep learning has attracted the researchers and the academicians in developing the medical imaging tools based on historical data. However, there are many challenges while blending the deep learning in medical imaging. Some of the key areas where deep learning seems to be challenging are: massive datasets for training, overfitting of data, variations in input, quality of data, understanding of the context, high volume of data and many more. This paper focuses on the deep learning challenges with biomedical imaging. The techniques for analyzing medical images in a heterogeneous environment are also discussed.