| Author: | Ho, Siu Cheong |
| Title: | Tongue diagnosis of TCM via advanced computer vision techniques enhancing remote TCM tongue diagnosis : a mixed-methods study on ai-assisted tongue feature recognition and regional analysis |
| Advisors: | Qin, Harry (SN) |
| Degree: | DHSc |
| Year: | 2025 |
| Department: | Faculty of Health and Social Sciences |
| Pages: | xxi, 193 pages : color illustrations |
| Language: | English |
| Abstract: | Background: The integration of Traditional Chinese Medicine (TCM) into telemedicine has encountered challenges related to the standardization of image quality and diagnostic accuracy in remote tongue diagnosis. This study seeks to address these issues by developing and validating AI systems designed to enhance the reliability and effectiveness of remote TCM consultations. Methods: Employing a sequential mixed-methods design, the research commenced with focus group interviews involving experienced TCM practitioners to gather insights on the practical challenges of remote diagnosis. These findings informed the creation of two specialized AI systems: SignNet for holistic tongue feature recognition, trained on 5,109 mobile-captured tongue images, and OrganNet for regional analysis, utilizing 4,645 images with organ-specific annotations. Both systems incorporate advanced Convolutional Neural Networks (CNN) architectures tailored to align with traditional TCM diagnostic principles. Results: The qualitative phase identified significant challenges in remote diagnosis, including image variability, technical issues, and concerns over diagnostic accuracy. SignNet outperformed existing models by approximately 4% in F1-score, while OrganNet achieved an average accuracy of 81.34% in assessing organ conditions, closely mirroring the performance of human experts. Notably, OrganNet excelled in assessing lung (72.29% accuracy) and kidney (85.83% F1-score) conditions, demonstrating the potential of AI to standardize and enhance remote TCM diagnostics. Conclusions: This research contributes significantly to the field by introducing comprehensive datasets for real-world mobile-captured tongue images, proposing innovative CNN-based AI systems that align with TCM principles, and establishing a framework for AI integration into TCM practice. While the AI systems showed promise in enhancing diagnostic consistency and accuracy, challenges remain, particularly in detecting subtle tongue features under variable lighting conditions. Future research should focus on improving feature detection, integrating with other TCM diagnostic methods, and expanding the application to broader TCM diagnostic areas. This study paves the way for more reliable and accessible remote TCM consultations, potentially revolutionizing telemedicine in traditional medicine. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 8647.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 2.53 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14195