Deep learning in generating radiology reports: A survey

This investigation focuses on the following critical challenges: understanding radiology text/image structures and datasets, applying DL algorithms (mainly CNN and RNN), generating Radiology text, and improving existing DL based models and evaluation metrics.