Estimation of quality of life using sublingual vein images based on deep learning and traditional Chinese medicine

A ring light and a polarizing filter are attached to a smartphone to capture cross-polarized images of sublingual veins. This setup minimizes the glare caused by saliva and allows for clear visualization of deep sublingual veins, which have long served as notable diagnostic indicators in Traditional Chinese Medicine. In this study, quality of life is evaluated using the Chinese Medicine Quality of Life-11 Dimensions Assessment scale, and deep learning is applied based on the ResNet50 architecture. Once trained, the neural network can predict quality of life from new sublingual vein images.