An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer
Abstract To retrospectively develop and validate an interpretable deep learning model and nomogram utilizing endoscopic ultrasound (EUS) images to predict pancreatic neuroendocrine tumors (PNETs). Following confirmation via pathological examination, a retrospective analysis was performed on a cohort...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-84749-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571718460243968 |
|---|---|
| author | Nan Yi Shuangyang Mo Yan Zhang Qi Jiang Yingwei Wang Cheng Huang Shanyu Qin Haixing Jiang |
| author_facet | Nan Yi Shuangyang Mo Yan Zhang Qi Jiang Yingwei Wang Cheng Huang Shanyu Qin Haixing Jiang |
| author_sort | Nan Yi |
| collection | DOAJ |
| description | Abstract To retrospectively develop and validate an interpretable deep learning model and nomogram utilizing endoscopic ultrasound (EUS) images to predict pancreatic neuroendocrine tumors (PNETs). Following confirmation via pathological examination, a retrospective analysis was performed on a cohort of 266 patients, comprising 115 individuals diagnosed with PNETs and 151 with pancreatic cancer. These patients were randomly assigned to the training or test group in a 7:3 ratio. The least absolute shrinkage and selection operator algorithm was employed to reduce the dimensionality of deep learning (DL) features extracted from pre-standardized EUS images. The retained nonzero coefficient features were subsequently applied to develop predictive eight DL models based on distinct machine learning algorithms. The optimal DL model was identified and used to establish a clinical signature, which subsequently informed the construction and evaluation of a nomogram. Gradient-weighted Class Activation Mapping (Grad-CAM) and Shapley Additive Explanations (SHAP) were implemented to interpret and visualize the model outputs. A total of 2048 DL features were initially extracted, from which only 27 features with coefficients greater than zero were retained. The support vector machine (SVM) DL model demonstrated exceptional performance, achieving area under the curve (AUC) values of 0.948 and 0.795 in the training and test groups, respectively. Additionally, a nomogram was developed, incorporating both DL and clinical signatures, and was visually represented for practical application. Finally, the calibration curves, decision curve analysis (DCA) plots, and clinical impact curves (CIC) exhibited by the DL model and nomogram indicated high accuracy. The application of Grad-CAM and SHAP enhanced the interpretability of these models. These methodologies contributed substantial net benefits to clinical decision-making processes. A novel interpretable DL model and nomogram were developed and validated using EUS images, cooperating with machine learning algorithms. This approach demonstrates significant potential for enhancing the clinical applicability of EUS in predicting PNETs from pancreatic cancer, thereby offering valuable insights for future research and implementation. |
| format | Article |
| id | doaj-art-495347ecca0549d4ba3f1e2393802d1e |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-495347ecca0549d4ba3f1e2393802d1e2025-02-02T12:22:29ZengNature PortfolioScientific Reports2045-23222025-01-0115111910.1038/s41598-024-84749-7An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancerNan Yi0Shuangyang Mo1Yan Zhang2Qi Jiang3Yingwei Wang4Cheng Huang5Shanyu Qin6Haixing Jiang7Gastroenterology Department, The First Affiliated Hospital of Guangxi Medical UniversityGastroenterology Department, The First Affiliated Hospital of Guangxi Medical UniversityThe First Affiliated Hospital of Shandong First Medical University & Shandong Provincial Qianfoshan HospitalGastroenterology Department, The First Affiliated Hospital of Guangxi Medical UniversityLiuzhou People’s Hospital Affiliated to Guangxi Medical UniversityLiuzhou People’s Hospital Affiliated to Guangxi Medical UniversityGastroenterology Department, The First Affiliated Hospital of Guangxi Medical UniversityGastroenterology Department, The First Affiliated Hospital of Guangxi Medical UniversityAbstract To retrospectively develop and validate an interpretable deep learning model and nomogram utilizing endoscopic ultrasound (EUS) images to predict pancreatic neuroendocrine tumors (PNETs). Following confirmation via pathological examination, a retrospective analysis was performed on a cohort of 266 patients, comprising 115 individuals diagnosed with PNETs and 151 with pancreatic cancer. These patients were randomly assigned to the training or test group in a 7:3 ratio. The least absolute shrinkage and selection operator algorithm was employed to reduce the dimensionality of deep learning (DL) features extracted from pre-standardized EUS images. The retained nonzero coefficient features were subsequently applied to develop predictive eight DL models based on distinct machine learning algorithms. The optimal DL model was identified and used to establish a clinical signature, which subsequently informed the construction and evaluation of a nomogram. Gradient-weighted Class Activation Mapping (Grad-CAM) and Shapley Additive Explanations (SHAP) were implemented to interpret and visualize the model outputs. A total of 2048 DL features were initially extracted, from which only 27 features with coefficients greater than zero were retained. The support vector machine (SVM) DL model demonstrated exceptional performance, achieving area under the curve (AUC) values of 0.948 and 0.795 in the training and test groups, respectively. Additionally, a nomogram was developed, incorporating both DL and clinical signatures, and was visually represented for practical application. Finally, the calibration curves, decision curve analysis (DCA) plots, and clinical impact curves (CIC) exhibited by the DL model and nomogram indicated high accuracy. The application of Grad-CAM and SHAP enhanced the interpretability of these models. These methodologies contributed substantial net benefits to clinical decision-making processes. A novel interpretable DL model and nomogram were developed and validated using EUS images, cooperating with machine learning algorithms. This approach demonstrates significant potential for enhancing the clinical applicability of EUS in predicting PNETs from pancreatic cancer, thereby offering valuable insights for future research and implementation.https://doi.org/10.1038/s41598-024-84749-7Pancreatic neuroendocrine tumorsEndoscopic ultrasoundDeep learningMachine learningShapley Additive explanationsGradient-weighted class activation mapping |
| spellingShingle | Nan Yi Shuangyang Mo Yan Zhang Qi Jiang Yingwei Wang Cheng Huang Shanyu Qin Haixing Jiang An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer Scientific Reports Pancreatic neuroendocrine tumors Endoscopic ultrasound Deep learning Machine learning Shapley Additive explanations Gradient-weighted class activation mapping |
| title | An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer |
| title_full | An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer |
| title_fullStr | An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer |
| title_full_unstemmed | An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer |
| title_short | An endoscopic ultrasound-based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer |
| title_sort | endoscopic ultrasound based interpretable deep learning model and nomogram for distinguishing pancreatic neuroendocrine tumors from pancreatic cancer |
| topic | Pancreatic neuroendocrine tumors Endoscopic ultrasound Deep learning Machine learning Shapley Additive explanations Gradient-weighted class activation mapping |
| url | https://doi.org/10.1038/s41598-024-84749-7 |
| work_keys_str_mv | AT nanyi anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT shuangyangmo anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT yanzhang anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT qijiang anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT yingweiwang anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT chenghuang anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT shanyuqin anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT haixingjiang anendoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT nanyi endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT shuangyangmo endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT yanzhang endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT qijiang endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT yingweiwang endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT chenghuang endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT shanyuqin endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer AT haixingjiang endoscopicultrasoundbasedinterpretabledeeplearningmodelandnomogramfordistinguishingpancreaticneuroendocrinetumorsfrompancreaticcancer |