대한의용생체공학회:의공학회지 (Journal of Biomedical Engineering Research)

  • 제45권2호
  • /
  • Pages.95-100
  • /
  • 2024
  • /
  • 1229-0807(pISSN)
  • /
  • 2288-9396(eISSN)
대한의용생체공학회 (The Korean Society of Medical and Biological Engineering)
권호 표지
DOI QR코드

DOI QR Code

안면 백반증 치료 평가를 위한 딥러닝 기반 자동화 분석 시스템 개발

Development of a Deep Learning-Based Automated Analysis System for Facial Vitiligo Treatment Evaluation

  • 이세나 (연세대학교 원주의과대학 정밀의학과) ;
  • 허연우 (연세대학교 원주의과대학 피부과학교실) ;
  • 이솔암 (연세대학교 원주의과대학 피부과학교실) ;
  • 박성빈 (연세대학교 원주의과대학 정밀의학과)
  • Sena Lee (Department of Precision Medicine, Yonsei University Wonju College of Medicine) ;
  • Yeon-Woo Heo (Department of Dermatology, Yonsei University Wonju College of Medicine) ;
  • Solam Lee (Department of Dermatology, Yonsei University Wonju College of Medicine) ;
  • Sung Bin Park (Department of Precision Medicine, Yonsei University Wonju College of Medicine)
  • 투고 : 2024.04.11
  • 심사 : 2024.04.24
  • 발행 : 2024.04.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

Vitiligo is a condition characterized by the destruction or dysfunction of melanin-producing cells in the skin, resulting in a loss of skin pigmentation. Facial vitiligo, specifically affecting the face, significantly impacts patients' appearance, thereby diminishing their quality of life. Evaluating the efficacy of facial vitiligo treatment typically relies on subjective assessments, such as the Facial Vitiligo Area Scoring Index (F-VASI), which can be time-consuming and subjective due to its reliance on clinical observations like lesion shape and distribution. Various machine learning and deep learning methods have been proposed for segmenting vitiligo areas in facial images, showing promising results. However, these methods often struggle to accurately segment vitiligo lesions irregularly distributed across the face. Therefore, our study introduces a framework aimed at improving the segmentation of vitiligo lesions on the face and providing an evaluation of vitiligo lesions. Our framework for facial vitiligo segmentation and lesion evaluation consists of three main steps. Firstly, we perform face detection to minimize background areas and identify the face area of interest using high-quality ultraviolet photographs. Secondly, we extract facial area masks and vitiligo lesion masks using a semantic segmentation network-based approach with the generated dataset. Thirdly, we automatically calculate the vitiligo area relative to the facial area. We evaluated the performance of facial and vitiligo lesion segmentation using an independent test dataset that was not included in the training and validation, showing excellent results. The framework proposed in this study can serve as a useful tool for evaluating the diagnosis and treatment efficacy of vitiligo.

키워드

과제정보

본 연구는 과학기술정보통신부(Ministry of Science and ICT, MSIT)의 재원으로 한국연구재단(National Research Foundation of Korea, NRF) 과제의 지원을 받아 수행하였음 (2022R1A2C2091160). 본 연구는 산업통상자원부와 한국산업기술진흥원의 "지역혁신클러스터육성 (R&D, P0025442)"사업의 지원을 받아 수행된 연구결과임.

참고문헌

  1. Chan MF, Thng TGS, Aw CWD, Goh BK, Lee SM, Chua TL. Investigating factors associated with quality of life of vitiligo patients in singapore. International journal of nursing practice. 2013;19:3-10. 
  2. Bae JM, Jung YS, Jung HM, Park JH, Hann SK. Classification of facial vitiligo: A cluster analysis of 473 patients. Pigment Cell & Melanoma Research. 2018;31(5):585-91. 
  3. Hamzavi I, Jain H, McLean D, Shapiro J, Zeng H, Lui H. Parametric modeling of narrowband UV-B phototherapy for vitiligo using a novel quantitative tool: the Vitiligo Area Scoring Index. Archives of Dermatology. 2004;140(6):677-83. 
  4. Bae JM, Zubair R, Ju HJ, Kohli I, Lee HN, Eun SH. Development and validation of the fingertip unit for assessing Facial Vitiligo Area Scoring Index. Journal of the American Academy of Dermatology. 2022;86(2):387-93. 
  5. Low M, Huang V, Raina P. Automating vitiligo skin lesion segmentation using convolutional neural networks. International journal of nursing practice. 2020;1-4. 
  6. Guo L, Yang Y, Ding H, Zheng H, Yang H, Xie J. A deep learning-based hybrid artificial intelligence model for the detection and severity assessment of vitiligo lesions. International journal of nursing practice. 2020;10(10). 
  7. Sharma S, Guleria K, Kumar S, Tiwari S. Deep Learning based Model for Detection of Vitiligo Skin Disease using Pretrained Inception V3. International Journal of Mathematical, Engineering and Management Sciences. 2023;8(5):1024 
  8. Khatibi T, Rezaei N, Ataei Fashtami L, Totonchi M. Proposing a novel unsupervised stack ensemble of deep and conventional image segmentation (SEDCIS) method for localizing vitiligo lesions in skin images. Skin Research and Technology. 2021;27(2):126-37. 
  9. Hillmer D, Merhi R, Boniface K, Taieb A, Barnetche T, Seneschal J. Evaluation of facial vitiligo severity with a mixed clinical and artificial intelligence approach.Journal of Investigative Dermatology. 2024;144(2):351-7. 
  10. Neri P, Fiaschi M, Menchini G. Semi-Automatic tool for vitiligo detection and analysis. Journal of imaging. 2020;6(3):14. 
  11. Deng J, Guo J, Ververas E, Kotsia I, Zafeiriou S. Retinaface: Single-shot multi-level face localisation in the wild. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020;5203-12. 
  12. Lee, GP., Kim, YJ., Lee, S., Kim, KG. Classification of anteroposterior/lateral images and segmentation of the radius using deep learning in wrist X-rays images. Journal of Biomedical Engineering Research. 2020;41(2):94-100. 
  13. Sun K, Xiao B, Liu D, Wang J. Deep high-resolution representation learning for human pose estimation. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2019;5693-703. 
  14. Liu J, Yao Y, Hou W, Cui M, Xie X, Zhang C. Boosting semantic human matting with coarse annotations. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020;8563-72. 
  15. Hong Y, Pan H, Sun W, Jia Y. Deep dual-resolution networks for real-time and accurate semantic segmentation of road scenes. arXiv preprint arXiv. 2021; 210106085. 
  16. Shrivastava A, Gupta A, Girshick R. Training region-based object detectors with online hard example mining. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016;761-9. 
  17. Yanling, L.I., Kong, AWK., Thng, S. Segmenting vitiligo on clinical face images using CNN trained on synthetic and internet images. IEEE Journal of Biomedical and Health Informatics. 2021 25(8):3082-3093. 
  18. Marin Dit Bertoud, Quentin. Reliability and agreement testing of a new automated measurement method to determine facial vitiligo extent using standardized ultraviolet images and a dedicated algorithm. British Journal of Dermatology. 2024;190(1):62-69 
  19. Hu, G., Zheng, Y., Yan, H., Hua, G., Yan, Y. Mask-guided cycle-GAN for specular highlight removal. Pattern Recognition Letters. 2022;161:108-114. 
  20. Su, T., Zhou, Y., Yu, Y., Du, S. Highlight Removal of Multi-View Facial Images. Sensors. 2022; 22(17):6656. 
  21. Guo, L., Yang, Y., Ding, H., Zheng, H., Yang, H., Xie, J., Ge, Y. A deep learning-based hybrid artificial intelligence model for the detection and severity assessment of vitiligo lesions. Annals of Translational Medicine. 2022;10(10). 
  22. Toh, JJH., Bhoi, S., Tan, VWD., Chuah, SY., Jhingan, A., Kong, AWK., Thng, STG. Automated scoring of vitiligo using superpixel-generated computerized digital image analysis of clinical photographs: a novel and consistent way to score vitiligo. British Journal of Dermatology. 2018;179(1):220-221.