Weakly supervised multiple-instance active learning for tooth-marked tongue recognition
IntroductionRecognizing a tooth-marked tongue has important clinical diagnostic value in traditional Chinese medicine. Current deep learning methods for tooth mark detection require extensive manual labeling and tongue segmentation, which is labor-intensive. Therefore, we propose a weakly supervised...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-06-01
|
| Series: | Frontiers in Physiology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fphys.2025.1598850/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850212264722825216 |
|---|---|
| author | Feilin Deng Shangxuan Li Zizhu Yang Wu Zhou |
| author_facet | Feilin Deng Shangxuan Li Zizhu Yang Wu Zhou |
| author_sort | Feilin Deng |
| collection | DOAJ |
| description | IntroductionRecognizing a tooth-marked tongue has important clinical diagnostic value in traditional Chinese medicine. Current deep learning methods for tooth mark detection require extensive manual labeling and tongue segmentation, which is labor-intensive. Therefore, we propose a weakly supervised multipleinstance active learning model for tooth-marked tongue recognition, aiming to eliminate preprocessing segmentation and reduce the annotation workload while maintaining diagnostic accuracy.MethodWe propose a one-stage method tongenerate tooth mark instances that eliminates the need for pre-segmentation of the tongue. To make full use of unlabeled data, we introduce a semisupervised learning paradigm to pseudo-label unlabeled tongue images with high model confidence in active learning and incorporate them into the training set to improve the training efficiency of the active learning model. In addition, we propose an instance-level hybrid query method considering the diversity of tooth marks.ResultExperimental results on clinical tongue images verify the effectiveness of the proposed method, which achieves an accuracy of 93.88% for tooth-marked tongue recognition, outperforming the recently introduced weakly supervised approaches.ConclusionThe proposed method is effective with only a small amount of image-level annotation, and its performance is comparable to that of image-level annotation, instance-level annotation and pixel-level annotation, which require a large number of tooth markers. Our method significantly reduces the annotation cost of the binary classification task of traditional Chinese medicine tooth mark recognition. |
| format | Article |
| id | doaj-art-d329252656fe4dafa05635e49ed0889c |
| institution | OA Journals |
| issn | 1664-042X |
| language | English |
| publishDate | 2025-06-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Physiology |
| spelling | doaj-art-d329252656fe4dafa05635e49ed0889c2025-08-20T02:09:23ZengFrontiers Media S.A.Frontiers in Physiology1664-042X2025-06-011610.3389/fphys.2025.15988501598850Weakly supervised multiple-instance active learning for tooth-marked tongue recognitionFeilin DengShangxuan LiZizhu YangWu ZhouIntroductionRecognizing a tooth-marked tongue has important clinical diagnostic value in traditional Chinese medicine. Current deep learning methods for tooth mark detection require extensive manual labeling and tongue segmentation, which is labor-intensive. Therefore, we propose a weakly supervised multipleinstance active learning model for tooth-marked tongue recognition, aiming to eliminate preprocessing segmentation and reduce the annotation workload while maintaining diagnostic accuracy.MethodWe propose a one-stage method tongenerate tooth mark instances that eliminates the need for pre-segmentation of the tongue. To make full use of unlabeled data, we introduce a semisupervised learning paradigm to pseudo-label unlabeled tongue images with high model confidence in active learning and incorporate them into the training set to improve the training efficiency of the active learning model. In addition, we propose an instance-level hybrid query method considering the diversity of tooth marks.ResultExperimental results on clinical tongue images verify the effectiveness of the proposed method, which achieves an accuracy of 93.88% for tooth-marked tongue recognition, outperforming the recently introduced weakly supervised approaches.ConclusionThe proposed method is effective with only a small amount of image-level annotation, and its performance is comparable to that of image-level annotation, instance-level annotation and pixel-level annotation, which require a large number of tooth markers. Our method significantly reduces the annotation cost of the binary classification task of traditional Chinese medicine tooth mark recognition.https://www.frontiersin.org/articles/10.3389/fphys.2025.1598850/fulltooth-marked tongueweakly supervised learningmultiple instance learningactive learningpseudo label |
| spellingShingle | Feilin Deng Shangxuan Li Zizhu Yang Wu Zhou Weakly supervised multiple-instance active learning for tooth-marked tongue recognition Frontiers in Physiology tooth-marked tongue weakly supervised learning multiple instance learning active learning pseudo label |
| title | Weakly supervised multiple-instance active learning for tooth-marked tongue recognition |
| title_full | Weakly supervised multiple-instance active learning for tooth-marked tongue recognition |
| title_fullStr | Weakly supervised multiple-instance active learning for tooth-marked tongue recognition |
| title_full_unstemmed | Weakly supervised multiple-instance active learning for tooth-marked tongue recognition |
| title_short | Weakly supervised multiple-instance active learning for tooth-marked tongue recognition |
| title_sort | weakly supervised multiple instance active learning for tooth marked tongue recognition |
| topic | tooth-marked tongue weakly supervised learning multiple instance learning active learning pseudo label |
| url | https://www.frontiersin.org/articles/10.3389/fphys.2025.1598850/full |
| work_keys_str_mv | AT feilindeng weaklysupervisedmultipleinstanceactivelearningfortoothmarkedtonguerecognition AT shangxuanli weaklysupervisedmultipleinstanceactivelearningfortoothmarkedtonguerecognition AT zizhuyang weaklysupervisedmultipleinstanceactivelearningfortoothmarkedtonguerecognition AT wuzhou weaklysupervisedmultipleinstanceactivelearningfortoothmarkedtonguerecognition |