Recent Progress of Medical CT Image Processing Based on Deep Learning

Abstract

Computed tomography (CT) can provide comprehensive and high-resolution sectional anatomy images of human tissues and has been widely used in medical examination and diagnosis. With the rapid development of machine learning, medical CT image processing based on deep learning has become a hotspot and achieved great progress. In this paper, the characteristics of CT image are analyzed. Then the research status and progress of medical CT image processing based on deep learning are investigated. The common deep learning models and typical applications are reviewed, which include segmentation, classification and detection. Medical CT image processing based on deep learning has shown higher performance compared with traditional methods, which can reduce the burden of doctors and assist doctors to make more accurate decisions. However, it still faces the challenges of insufficient large-scale datasets, difficult dataset annotation and interpretability of the models that applications need. In the conclusion, the future development trend is analyzed.