Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution
Introduction: Deep learning models can assess the quality of images and discriminate among abnormalities in small bowel capsule endoscopy (CE), reducing fatigue and the time needed for diagnosis. They serve as a decision support system, partially automating the diagnosis process by providing probabi...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wolters Kluwer – Medknow Publications
2024-03-01
|
| Series: | Singapore Medical Journal |
| Subjects: | |
| Online Access: | https://journals.lww.com/10.4103/singaporemedj.SMJ-2023-187 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823863778444312576 |
|---|---|
| author | Bochao Jiang Michael Dorosan Justin Wen Hao Leong Marcus Eng Hock Ong Sean Shao Wei Lam Tiing Leong Ang |
| author_facet | Bochao Jiang Michael Dorosan Justin Wen Hao Leong Marcus Eng Hock Ong Sean Shao Wei Lam Tiing Leong Ang |
| author_sort | Bochao Jiang |
| collection | DOAJ |
| description | Introduction:
Deep learning models can assess the quality of images and discriminate among abnormalities in small bowel capsule endoscopy (CE), reducing fatigue and the time needed for diagnosis. They serve as a decision support system, partially automating the diagnosis process by providing probability predictions for abnormalities.
Methods:
We demonstrated the use of deep learning models in CE image analysis, specifically by piloting a bowel preparation model (BPM) and an abnormality detection model (ADM) to determine frame-level view quality and the presence of abnormal findings, respectively. We used convolutional neural network-based models pretrained on large-scale open-domain data to extract spatial features of CE images that were then used in a dense feed-forward neural network classifier. We then combined the open-source Kvasir-Capsule dataset (n = 43) and locally collected CE data (n = 29).
Results:
Model performance was compared using averaged five-fold and two-fold cross-validation for BPMs and ADMs, respectively. The best BPM model based on a pre-trained ResNet50 architecture had an area under the receiver operating characteristic and precision-recall curves of 0.969±0.008 and 0.843±0.041, respectively. The best ADM model, also based on ResNet50, had top-1 and top-2 accuracies of 84.03±0.051 and 94.78±0.028, respectively. The models could process approximately 200–250 images per second and showed good discrimination on time-critical abnormalities such as bleeding.
Conclusion:
Our pilot models showed the potential to improve time to diagnosis in CE workflows. To our knowledge, our approach is unique to the Singapore context. The value of our work can be further evaluated in a pragmatic manner that is sensitive to existing clinician workflow and resource constraints. |
| format | Article |
| id | doaj-art-98e43a5cc0714edbbb8602ba3c8a13e6 |
| institution | Kabale University |
| issn | 0037-5675 2737-5935 |
| language | English |
| publishDate | 2024-03-01 |
| publisher | Wolters Kluwer – Medknow Publications |
| record_format | Article |
| series | Singapore Medical Journal |
| spelling | doaj-art-98e43a5cc0714edbbb8602ba3c8a13e62025-02-09T10:19:44ZengWolters Kluwer – Medknow PublicationsSingapore Medical Journal0037-56752737-59352024-03-0165313314010.4103/singaporemedj.SMJ-2023-187Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institutionBochao JiangMichael DorosanJustin Wen Hao LeongMarcus Eng Hock OngSean Shao Wei LamTiing Leong AngIntroduction: Deep learning models can assess the quality of images and discriminate among abnormalities in small bowel capsule endoscopy (CE), reducing fatigue and the time needed for diagnosis. They serve as a decision support system, partially automating the diagnosis process by providing probability predictions for abnormalities. Methods: We demonstrated the use of deep learning models in CE image analysis, specifically by piloting a bowel preparation model (BPM) and an abnormality detection model (ADM) to determine frame-level view quality and the presence of abnormal findings, respectively. We used convolutional neural network-based models pretrained on large-scale open-domain data to extract spatial features of CE images that were then used in a dense feed-forward neural network classifier. We then combined the open-source Kvasir-Capsule dataset (n = 43) and locally collected CE data (n = 29). Results: Model performance was compared using averaged five-fold and two-fold cross-validation for BPMs and ADMs, respectively. The best BPM model based on a pre-trained ResNet50 architecture had an area under the receiver operating characteristic and precision-recall curves of 0.969±0.008 and 0.843±0.041, respectively. The best ADM model, also based on ResNet50, had top-1 and top-2 accuracies of 84.03±0.051 and 94.78±0.028, respectively. The models could process approximately 200–250 images per second and showed good discrimination on time-critical abnormalities such as bleeding. Conclusion: Our pilot models showed the potential to improve time to diagnosis in CE workflows. To our knowledge, our approach is unique to the Singapore context. The value of our work can be further evaluated in a pragmatic manner that is sensitive to existing clinician workflow and resource constraints.https://journals.lww.com/10.4103/singaporemedj.SMJ-2023-187capsule endoscopydetectiondiagnosismachine learning |
| spellingShingle | Bochao Jiang Michael Dorosan Justin Wen Hao Leong Marcus Eng Hock Ong Sean Shao Wei Lam Tiing Leong Ang Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution Singapore Medical Journal capsule endoscopy detection diagnosis machine learning |
| title | Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution |
| title_full | Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution |
| title_fullStr | Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution |
| title_full_unstemmed | Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution |
| title_short | Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution |
| title_sort | development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy a pilot study in a singapore institution |
| topic | capsule endoscopy detection diagnosis machine learning |
| url | https://journals.lww.com/10.4103/singaporemedj.SMJ-2023-187 |
| work_keys_str_mv | AT bochaojiang developmentandvalidationofadeeplearningsystemfordetectionofsmallbowelpathologiesincapsuleendoscopyapilotstudyinasingaporeinstitution AT michaeldorosan developmentandvalidationofadeeplearningsystemfordetectionofsmallbowelpathologiesincapsuleendoscopyapilotstudyinasingaporeinstitution AT justinwenhaoleong developmentandvalidationofadeeplearningsystemfordetectionofsmallbowelpathologiesincapsuleendoscopyapilotstudyinasingaporeinstitution AT marcusenghockong developmentandvalidationofadeeplearningsystemfordetectionofsmallbowelpathologiesincapsuleendoscopyapilotstudyinasingaporeinstitution AT seanshaoweilam developmentandvalidationofadeeplearningsystemfordetectionofsmallbowelpathologiesincapsuleendoscopyapilotstudyinasingaporeinstitution AT tiingleongang developmentandvalidationofadeeplearningsystemfordetectionofsmallbowelpathologiesincapsuleendoscopyapilotstudyinasingaporeinstitution |