Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images
Abstract There is a currently an unmet need for non-invasive methods to predict the risk of esophageal squamous cell carcinoma (ESCC). Previously, we found that specific soft palate morphologies are strongly associated with increased ESCC risk. However, there is currently no artificial intelligence...
Saved in:
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-86829-8 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571756467978240 |
|---|---|
| author | Kotaro Waki Katsuya Nagaoka Keishi Okubo Masato Kiyama Ryosuke Gushima Kento Ohno Munenori Honda Akira Yamasaki Kenshi Matsuno Yoki Furuta Hideaki Miyamoto Hideaki Naoe Motoki Amagasaki Yasuhito Tanaka |
| author_facet | Kotaro Waki Katsuya Nagaoka Keishi Okubo Masato Kiyama Ryosuke Gushima Kento Ohno Munenori Honda Akira Yamasaki Kenshi Matsuno Yoki Furuta Hideaki Miyamoto Hideaki Naoe Motoki Amagasaki Yasuhito Tanaka |
| author_sort | Kotaro Waki |
| collection | DOAJ |
| description | Abstract There is a currently an unmet need for non-invasive methods to predict the risk of esophageal squamous cell carcinoma (ESCC). Previously, we found that specific soft palate morphologies are strongly associated with increased ESCC risk. However, there is currently no artificial intelligence (AI) system that utilizes oral images for ESCC risk assessment. Here, we evaluated three AI models and three fine-tuning approaches with regard to their ESCC predictive power. Our dataset contained 539 cases, which were subdivided into 221 high-risk cases (2491 images) and 318 non-high-risk cases (2524 images). We used 480 cases (4295 images) for the training dataset, and the rest for validation. The Bilinear convolutional neural network (CNN) model (especially when pre-trained on fractal images) demonstrated diagnostic precision that was comparable to or better than other models for distinguishing between high-risk and non-high-risk groups. In addition, when tested with a small number of images containing soft palate data, the model showed high precision: the best AUC model had 0.91 (sensitivity 0.86, specificity 0.79). This study presents a significant advance in the development of an AI-based non-invasive screening tool for the identification of high-risk ESCC patients. The approach may be particularly suitable for institutes with limited medical imaging resources. |
| format | Article |
| id | doaj-art-d04f0d93aadb4a27a989d6e539d177d4 |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-d04f0d93aadb4a27a989d6e539d177d42025-02-02T12:23:52ZengNature PortfolioScientific Reports2045-23222025-02-0115111110.1038/s41598-025-86829-8Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate imagesKotaro Waki0Katsuya Nagaoka1Keishi Okubo2Masato Kiyama3Ryosuke Gushima4Kento Ohno5Munenori Honda6Akira Yamasaki7Kenshi Matsuno8Yoki Furuta9Hideaki Miyamoto10Hideaki Naoe11Motoki Amagasaki12Yasuhito Tanaka13Department of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityFaculty of Advanced Science and Technology, Kumamoto UniversityFaculty of Advanced Science and Technology, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityFaculty of Advanced Science and Technology, Kumamoto UniversityDepartment of Gastroenterology and Hepatology, Faculty of Life Sciences, Kumamoto UniversityAbstract There is a currently an unmet need for non-invasive methods to predict the risk of esophageal squamous cell carcinoma (ESCC). Previously, we found that specific soft palate morphologies are strongly associated with increased ESCC risk. However, there is currently no artificial intelligence (AI) system that utilizes oral images for ESCC risk assessment. Here, we evaluated three AI models and three fine-tuning approaches with regard to their ESCC predictive power. Our dataset contained 539 cases, which were subdivided into 221 high-risk cases (2491 images) and 318 non-high-risk cases (2524 images). We used 480 cases (4295 images) for the training dataset, and the rest for validation. The Bilinear convolutional neural network (CNN) model (especially when pre-trained on fractal images) demonstrated diagnostic precision that was comparable to or better than other models for distinguishing between high-risk and non-high-risk groups. In addition, when tested with a small number of images containing soft palate data, the model showed high precision: the best AUC model had 0.91 (sensitivity 0.86, specificity 0.79). This study presents a significant advance in the development of an AI-based non-invasive screening tool for the identification of high-risk ESCC patients. The approach may be particularly suitable for institutes with limited medical imaging resources.https://doi.org/10.1038/s41598-025-86829-8Esophageal squamous cell carcinomaArtificial intelligenceConvolutional neural networkSoft palateNon-invasive risk assessment |
| spellingShingle | Kotaro Waki Katsuya Nagaoka Keishi Okubo Masato Kiyama Ryosuke Gushima Kento Ohno Munenori Honda Akira Yamasaki Kenshi Matsuno Yoki Furuta Hideaki Miyamoto Hideaki Naoe Motoki Amagasaki Yasuhito Tanaka Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images Scientific Reports Esophageal squamous cell carcinoma Artificial intelligence Convolutional neural network Soft palate Non-invasive risk assessment |
| title | Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images |
| title_full | Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images |
| title_fullStr | Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images |
| title_full_unstemmed | Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images |
| title_short | Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images |
| title_sort | optimizing ai models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images |
| topic | Esophageal squamous cell carcinoma Artificial intelligence Convolutional neural network Soft palate Non-invasive risk assessment |
| url | https://doi.org/10.1038/s41598-025-86829-8 |
| work_keys_str_mv | AT kotarowaki optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT katsuyanagaoka optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT keishiokubo optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT masatokiyama optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT ryosukegushima optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT kentoohno optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT munenorihonda optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT akirayamasaki optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT kenshimatsuno optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT yokifuruta optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT hideakimiyamoto optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT hideakinaoe optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT motokiamagasaki optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages AT yasuhitotanaka optimizingaimodelstopredictesophagealsquamouscellcarcinomariskbyincorporatingsmalldatasetsofsoftpalateimages |