Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet
<italic>Goal:</italic> Augment a small, imbalanced, wound dataset by using semi-supervised learning with a secondary dataset. Then utilize the augmented wound dataset for deep learning-based wound assessment. <italic>Methods:</italic> The clinically-validated Photographic Wou...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Open Journal of Engineering in Medicine and Biology |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10050724/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849418039017078784 |
|---|---|
| author | Ziyang Liu Emmanuel Agu Peder Pedersen Clifford Lindsay Bengisu Tulu Diane Strong |
| author_facet | Ziyang Liu Emmanuel Agu Peder Pedersen Clifford Lindsay Bengisu Tulu Diane Strong |
| author_sort | Ziyang Liu |
| collection | DOAJ |
| description | <italic>Goal:</italic> Augment a small, imbalanced, wound dataset by using semi-supervised learning with a secondary dataset. Then utilize the augmented wound dataset for deep learning-based wound assessment. <italic>Methods:</italic> The clinically-validated Photographic Wound Assessment Tool (PWAT) scores eight wound attributes: Size, Depth, Necrotic Tissue Type, Necrotic Tissue Amount, Granulation Tissue type, Granulation Tissue Amount, Edges, Periulcer Skin Viability to comprehensively assess chronic wound images. A small corpus of 1639 wound images labeled with ground truth PWAT scores was used as reference. A Semi-Supervised learning and Progressive Multi-Granularity training mechanism were used to leverage a secondary corpus of 9870 unlabeled wound images. Wound scoring utilized the EfficientNet Convolutional Neural Network on the augmented wound corpus. <italic>Results:</italic> Our proposed Semi-Supervised PMG EfficientNet (SS-PMG-EfficientNet) approach estimated all 8 PWAT sub-scores with classification accuracies and F1 scores of about 90% on average, and outperformed a comprehensive list of baseline models and had a 7% improvement over the prior state-of-the-art (without data augmentation). We also demonstrate that synthetic wound image generation using Generative Adversarial Networks (GANs) did not improve wound assessment. <italic>Conclusions:</italic> Semi-supervised learning on unlabeled wound images in a secondary dataset achieved impressive performance for deep learning-based wound grading. |
| format | Article |
| id | doaj-art-0dd534e396e0400a8669feb4673feb1a |
| institution | Kabale University |
| issn | 2644-1276 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Open Journal of Engineering in Medicine and Biology |
| spelling | doaj-art-0dd534e396e0400a8669feb4673feb1a2025-08-20T03:32:33ZengIEEEIEEE Open Journal of Engineering in Medicine and Biology2644-12762024-01-01540442010.1109/OJEMB.2023.324830710050724Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNetZiyang Liu0https://orcid.org/0000-0001-5419-1250Emmanuel Agu1https://orcid.org/0000-0002-3361-4952Peder Pedersen2https://orcid.org/0000-0002-7917-7209Clifford Lindsay3Bengisu Tulu4https://orcid.org/0000-0001-7226-1830Diane Strong5https://orcid.org/0000-0002-1756-0464Computer Science Department, Worcester Polytechnic Institute, Worcester, MA, USAComputer Science Department, Worcester Polytechnic Institute, Worcester, MA, USAElectrical and Computer Engineering Department, Worcester Polytechnic Institute, Worcester, MA, USADepartment of Radiology, University of Massachusetts Medical School, Worcester, MA, USAFoisie Business School, Worcester Polytechnic Institute, Worcester, MA, USAFoisie Business School, Worcester Polytechnic Institute, Worcester, MA, USA<italic>Goal:</italic> Augment a small, imbalanced, wound dataset by using semi-supervised learning with a secondary dataset. Then utilize the augmented wound dataset for deep learning-based wound assessment. <italic>Methods:</italic> The clinically-validated Photographic Wound Assessment Tool (PWAT) scores eight wound attributes: Size, Depth, Necrotic Tissue Type, Necrotic Tissue Amount, Granulation Tissue type, Granulation Tissue Amount, Edges, Periulcer Skin Viability to comprehensively assess chronic wound images. A small corpus of 1639 wound images labeled with ground truth PWAT scores was used as reference. A Semi-Supervised learning and Progressive Multi-Granularity training mechanism were used to leverage a secondary corpus of 9870 unlabeled wound images. Wound scoring utilized the EfficientNet Convolutional Neural Network on the augmented wound corpus. <italic>Results:</italic> Our proposed Semi-Supervised PMG EfficientNet (SS-PMG-EfficientNet) approach estimated all 8 PWAT sub-scores with classification accuracies and F1 scores of about 90% on average, and outperformed a comprehensive list of baseline models and had a 7% improvement over the prior state-of-the-art (without data augmentation). We also demonstrate that synthetic wound image generation using Generative Adversarial Networks (GANs) did not improve wound assessment. <italic>Conclusions:</italic> Semi-supervised learning on unlabeled wound images in a secondary dataset achieved impressive performance for deep learning-based wound grading.https://ieeexplore.ieee.org/document/10050724/Chronic woundsdata imbalancedata augmentationneural networkssmartphone assessment |
| spellingShingle | Ziyang Liu Emmanuel Agu Peder Pedersen Clifford Lindsay Bengisu Tulu Diane Strong Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet IEEE Open Journal of Engineering in Medicine and Biology Chronic wounds data imbalance data augmentation neural networks smartphone assessment |
| title | Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet |
| title_full | Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet |
| title_fullStr | Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet |
| title_full_unstemmed | Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet |
| title_short | Chronic Wound Image Augmentation and Assessment Using Semi-Supervised Progressive Multi-Granularity EfficientNet |
| title_sort | chronic wound image augmentation and assessment using semi supervised progressive multi granularity efficientnet |
| topic | Chronic wounds data imbalance data augmentation neural networks smartphone assessment |
| url | https://ieeexplore.ieee.org/document/10050724/ |
| work_keys_str_mv | AT ziyangliu chronicwoundimageaugmentationandassessmentusingsemisupervisedprogressivemultigranularityefficientnet AT emmanuelagu chronicwoundimageaugmentationandassessmentusingsemisupervisedprogressivemultigranularityefficientnet AT pederpedersen chronicwoundimageaugmentationandassessmentusingsemisupervisedprogressivemultigranularityefficientnet AT cliffordlindsay chronicwoundimageaugmentationandassessmentusingsemisupervisedprogressivemultigranularityefficientnet AT bengisutulu chronicwoundimageaugmentationandassessmentusingsemisupervisedprogressivemultigranularityefficientnet AT dianestrong chronicwoundimageaugmentationandassessmentusingsemisupervisedprogressivemultigranularityefficientnet |