A Review of Traditional Algorithms and Deep Learning for Dental Segmentation

The processing of medical images plays an important role in the medical profession. The prediction of diseases, postoperative recovery, and the classification of lesions all depend heavily on the diagnosis after imaging. Some unique attributes of medical images also make them significantly different from ordinary image processing. Due to the sensitivity of medical images to data and the complexity of the amount of data, the introduction of deep learning relies on complex human learning behaviors to process more complex medical images. This article summarizes various optimized traditional segmentation methods and deep learning methods, and conducts extensive paper reviews based on the years of algorithm development, processing effects, and optimization. It also follows up on the recent development of image segmentation at this stage in real time, focusing on dental images.