설명가능한 인공지능을 활용한 안면 특징 분석 기반 사상체질 검출
Sasang Constitution Detection Based on Facial Feature Analysis Using Explainable Artificial Intelligence
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Objectives The aim was to develop a method for detecting Sasang constitution based on the ratio of facial landmarks and provide an objective and reliable tool for Sasang constitution classification. Methods Facial images, KS-15 scores, and certainty scores were collected from subjects identified by Korean Medicine Data Center. Facial ratio landmarks were detected, yielding 2279 facial ratio features. Tree-based models were trained to classify Sasang constitution, and Shapley Additive Explanations (SHAP) analysis was employed to identify important facial features. Additionally, Body Mass Index (BMI) and personality questionnaire were incorporated as supplementary information to enhance model performance. Results Using the Tree-based models, the accuracy for classifying Taeeum, Soeum, and Soyang constitutions was 81.90%, 90.49%, and 81.90% respectively. SHAP analysis revealed important facial features, while the inclusion of BMI and personality questionnaire improved model performance. This demonstrates that facial ratio-based Sasang constitution analysis yields effective and accurate classification results. Conclusions Facial ratio-based Sasang constitution analysis provides rapid and objective results compared to traditional methods. This approach holds promise for enhancing personalized medicine in Korean traditional medicine.
Abstract
Objectives The aim was to develop a method for detecting Sasang constitution based on the ratio of facial landmarks and provide an objective and reliable tool for Sasang constitution classification. Methods Facial images, KS-15 scores, and certainty scores were collected from subjects identified by Korean Medicine Data Center. Facial ratio landmarks were detected, yielding 2279 facial ratio features. Tree-based models were trained to classify Sasang constitution, and Shapley Additive Explanations (SHAP) analysis was employed to identify important facial features. Additionally, Body Mass Index (BMI) and personality questionnaire were incorporated as supplementary information to enhance model performance. Results Using the Tree-based models, the accuracy for classifying Taeeum, Soeum, and Soyang constitutions was 81.90%, 90.49%, and 81.90% respectively. SHAP analysis revealed important facial features, while the inclusion of BMI and personality questionnaire improved model performance. This demonstrates that facial ratio-based Sasang constitution analysis yields effective and accurate classification results. Conclusions Facial ratio-based Sasang constitution analysis provides rapid and objective results compared to traditional methods. This approach holds promise for enhancing personalized medicine in Korean traditional medicine.
참고문헌 (19)
- Department of Sasang Constitutional Medicine, College of Korean Medicine, "Sasang Medicine" (2006), Jibmundang
- Lee SW, "Current researches on the methods of diagnosing sasang constitution : an overview", Evidence-Based Complementary and Alternative Medicine, 6, pp.43~49 (2009)[DOI]
- Lee JH, "Survey study about Sasangin's characteristics of face, voice, skin and pulse diagnosis", Journal of Sasang Constitutional Medicine, 19(3), pp.126~143 (2007)
- Kang B, "Prompt-RAG:Pioneering Vector Embedding-Free Retrieval-Augmented Generation in Niche Domains", Exemplified by Korean Medicine (2024)[DOI]
- Sohn EH, "The study of Sasangin's face", Journal of Sasang Constitutional Medicine, 17(3), pp.55~68 (2005)
- Do JH, "Sasang Constitutional Diagnosis Technology Using Facial Information", Journal of institute of Electronics Engineers, 37(7), pp.53~61 (2010)
- Lee J, "Facial features analysis of Sasang typology using facial images : A study on Caucasian American and native Korean populations", European Journal of Integrative Medicine, 57, pp.102210~ (2023)[DOI]
- Ahn I, "Sasang Constitution Classification using Convolutional Neural Network on Facial Images", Journal of Sasang Constitutional Medicine, 34(3), pp.31~40 (2022)[DOI]
- Fa S, "Multi-scale spatial–temporal attention graph convolutional networks for driver fatigue detection", Journal of Visual Communication and Image Representation, 93, pp.103826~ (2023)[DOI]
- Yan L, "Multi-feature fusing local directional ternary pattern for facial expressions signal recognition based on video communication system", Alexandria Engineering Journal, 63, pp.307~320 (2023)[DOI]
- Kim JK, "Explainable artificial intelligence and wearable sensor-based gait analysis to identify patients with osteopenia and sarcopenia in daily life", Biosensors, 12(3), pp.167~ (2022)[DOI]
- Nazar M, "A systematic review of human–computer interaction and explainable artificial intelligence in healthcare with artificial intelligence techniques", IEEE Access, 9, pp.153316~153348 (2021)[DOI]
- Baek YH, "Development and validation of brief KS-15(Korea Sasang Constitutional Diagnostic Questionnaire)based on body shape, temperament and symptoms", Journal of Sasang Constitutional Medicine, 27(2), pp.211~221 (2015)[DOI]
- Establishing a Constitution Information Collection System for the Scientific Diagnosis of Constitutional Medicine, "Lee SW", Korea Institute of Oriental Medicine, ~ (2009)
- Ye RZ, "Effects of image quality on the accuracy human pose estimation and detection of eye lid opening", closing using OpenPose and DLib, 2022, pp.8(12), 330~
- Varela VP, "Looking at faces in the wild", Scientific Reports, 13(1), pp.783~ (2023)[DOI]
- Kim JK, "Detection of important features and comparison of datasets for fall detection based on wrist-wearable devices", Expert Systems with Applications, 234, pp.121034~ (2023)[DOI]
- Slack D, "Fooling lime and shap: Adversarial attacks on post hoc explanation methods", Proceedings of the AAAI, 2020, 180~186
- Lee SY, "Systematic review on researches of sasang constitution diagnosis using facial feature", Journal of Sasang Constitutional Medicine, 24(4), pp.17~27 (2012)