Research on Automatic Classification Methods for Tibetan Medicine Urine Diagnosis Images

作者：Gadeng Luosang, Duojie Renzeng, Sanzhi Cairang, Dongzhu Renqing, Nuo Qun, Nyima Tashi · 年份：2024 · DOI：10.1109/mvipit65697.2024.00009 · 研究领域：Traditional Chinese Medicine Studies、Medical Research and Treatments、Cell Image Analysis Techniques

Tibetan medicine, one of the four major traditional medical systems in the world, employs urine diagnosis as a unique and widely used method for disease identification in Tibetan medicine. This study developed a joint classification model that leverages deep neural networks to classify the features and attributes of images of urine used in Tibetan medicine. Tested on a dataset collected in real-world conditions, the model demonstrated a classification accuracy of up to 93.6%, laying a solid foundation for the future development of decision-support platforms specifically designed for urine diagnosis in Tibetan medicine.