Infrared Thermography-Based Predictive Model for Syndrome Differentiation of Chaihu Guizhi Ganjiang Decoction

Abstract

Objective: To evaluate the use of infrared thermography technology for objective and quantitative syndrome differentiation and treatment in traditional Chinese medicine (TCM), specifically in patients with Chaihu Guizhi Ganjiang Decoction syndrome. Methods: Data were collected from over 100 patients diagnosed with Chaihu Guizhi Ganjiang Decoction syndrome at Professor Li Leyu’s endocrinology clinic, Zhongshan Hospital of Traditional Chinese Medicine, Guangdong Province, between April 2021 and April 2022. Body surface temperature data were obtained using the MTI-EXPRO-2013-B infrared thermography system. Principal component analysis (PCA) was applied to differentiate temperature distribution characteristics between genders, and a neural network prediction model was constructed for syndrome diagnosis. Results: Infrared thermography effectively captured surface temperature characteristics of patients with Chaihu Guizhi Ganjiang Decoction syndrome. PCA identified one principal component with a variance explanation rate of 73.953% for females and two principal components with a cumulative variance explanation rate of 77.627% for males. The neural network model demonstrated high predictive performance, with an area under the ROC curve of 0.9743 for the training set and 0.9889 for the validation set. Sensitivity was 1, specificity 0.8636, precision 0.8846, accuracy 0.9333, and the F1 score 0.9388. Conclusion: Infrared thermography provides an innovative, objective, and quantitative method for syndrome differentiation and treatment in TCM. It represents a significant advancement in transitioning from traditional empirical approaches to modern, visualized, and precise diagnosis and treatment. This study underscores the potential of integrating advanced technologies in TCM for enhanced clinical application and modernization.