中医体质分类辨识技术的回顾和展望

A review and prospects of body constitution classification and identification techniques in traditional Chinese medicine

关键词:  中医体质  中医体质辨识  多模态  人工智能

Key words:  traditional Chinese medicine constitution  traditional Chinese medicine constitution identification  multimodality  artificialintelligence

摘要:

中医体质理论寻求建立以“治未病”为基础思想的疾病预防体系，体质辨识效果直接影响后续的健康干预等措施。目前中医体质的主要辨识方法为量表法和医生四诊法，缺乏标准性和客观性。对现有中医体质辨识技术与中医体质辅助辨识技术进行整理与回顾，并在现有研究基础上，结合以深度学习为主的人工智能技术和多模态数据融合思想，对未来中医体质分类辨识技术的发展方向作出探讨与展望。

Abstract:

The constitution theory of traditional Chinese medicine seeks to establish a disease prevention system based on the idea of "preventive treatment of diseases". The effectiveness of constitution identification directly affects the subsequent health interventions. At present,the main methods of traditional Chinese medicine constitution identification are the scale method and the four diagnostic methods of doctors,which lack standardization and objectivity. In this paper,we review the existing traditional Chinese medicine identification and assisted identification technology,and discuss the future development of traditional Chinese medicine constitution classification and identification techniques based on the existing research,combined with deep learning-based artificial intelligence technology and multimodal data fusion ideas.