Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma
Background: Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most aggressive cancers with an extremely poor prognosis. Radiomics has shown prognostic ability in multiple types of cancer including PDAC. However, the prognostic value of traditional radiomics pipelines, which are based on hand-cra...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2020-10-01
|
| Series: | Frontiers in Artificial Intelligence |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/article/10.3389/frai.2020.550890/full |
| _version_ | 1828401008939630592 |
|---|---|
| author | Yucheng Zhang Edrise M. Lobo-Mueller Paul Karanicolas Steven Gallinger Masoom A. Haider Masoom A. Haider Farzad Khalvati Farzad Khalvati Farzad Khalvati |
| author_facet | Yucheng Zhang Edrise M. Lobo-Mueller Paul Karanicolas Steven Gallinger Masoom A. Haider Masoom A. Haider Farzad Khalvati Farzad Khalvati Farzad Khalvati |
| author_sort | Yucheng Zhang |
| collection | DOAJ |
| description | Background: Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most aggressive cancers with an extremely poor prognosis. Radiomics has shown prognostic ability in multiple types of cancer including PDAC. However, the prognostic value of traditional radiomics pipelines, which are based on hand-crafted radiomic features alone is limited.Methods: Convolutional neural networks (CNNs) have been shown to outperform radiomics models in computer vision tasks. However, training a CNN from scratch requires a large sample size which is not feasible in most medical imaging studies. As an alternative solution, CNN-based transfer learning models have shown the potential for achieving reasonable performance using small datasets. In this work, we developed and validated a CNN-based transfer learning model for prognostication of overall survival in PDAC patients using two independent resectable PDAC cohorts.Results: The proposed transfer learning-based prognostication model for overall survival achieved the area under the receiver operating characteristic curve of 0.81 on the test cohort, which was significantly higher than that of the traditional radiomics model (0.54). To further assess the prognostic value of the models, the predicted probabilities of death generated from the two models were used as risk scores in a univariate Cox Proportional Hazard model and while the risk score from the traditional radiomics model was not associated with overall survival, the proposed transfer learning-based risk score had significant prognostic value with hazard ratio of 1.86 (95% Confidence Interval: 1.15–3.53, p-value: 0.04).Conclusions: This result suggests that transfer learning-based models may significantly improve prognostic performance in typical small sample size medical imaging studies. |
| first_indexed | 2024-12-10T09:41:09Z |
| format | Article |
| id | doaj.art-516f6b5d9a574f9998a87c77545a913e |
| institution | Directory Open Access Journal |
| issn | 2624-8212 |
| language | English |
| last_indexed | 2024-12-10T09:41:09Z |
| publishDate | 2020-10-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Artificial Intelligence |
| spelling | doaj.art-516f6b5d9a574f9998a87c77545a913e2022-12-22T01:54:00ZengFrontiers Media S.A.Frontiers in Artificial Intelligence2624-82122020-10-01310.3389/frai.2020.550890550890Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal AdenocarcinomaYucheng Zhang0Edrise M. Lobo-Mueller1Paul Karanicolas2Steven Gallinger3Masoom A. Haider4Masoom A. Haider5Farzad Khalvati6Farzad Khalvati7Farzad Khalvati8Department of Medical Imaging, University of Toronto, Toronto, ON, CanadaDepartment of Diagnostic Imaging and Department of Oncology, Faculty of Medicine and Dentistry, Cross Cancer Institute, University of Alberta, Edmonton, AB, CanadaDepartment of Surgery, Sunnybrook Health Sciences Centre, Toronto, ON, CanadaLunenfeld-Tanenbaum Research Institute, Sinai Health System, Toronto, ON, CanadaLunenfeld-Tanenbaum Research Institute, Sinai Health System, Toronto, ON, CanadaJoint Department of Medical Imaging, Sinai Health System, University Health Network, University of Toronto, Toronto, ON, CanadaDepartment of Medical Imaging, University of Toronto, Toronto, ON, CanadaResearch Institute, The Hospital for Sick Children, Toronto, ON, CanadaDepartment of Mechanical and Industrial Engineering, University of Toronto, Toronto, ON, CanadaBackground: Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most aggressive cancers with an extremely poor prognosis. Radiomics has shown prognostic ability in multiple types of cancer including PDAC. However, the prognostic value of traditional radiomics pipelines, which are based on hand-crafted radiomic features alone is limited.Methods: Convolutional neural networks (CNNs) have been shown to outperform radiomics models in computer vision tasks. However, training a CNN from scratch requires a large sample size which is not feasible in most medical imaging studies. As an alternative solution, CNN-based transfer learning models have shown the potential for achieving reasonable performance using small datasets. In this work, we developed and validated a CNN-based transfer learning model for prognostication of overall survival in PDAC patients using two independent resectable PDAC cohorts.Results: The proposed transfer learning-based prognostication model for overall survival achieved the area under the receiver operating characteristic curve of 0.81 on the test cohort, which was significantly higher than that of the traditional radiomics model (0.54). To further assess the prognostic value of the models, the predicted probabilities of death generated from the two models were used as risk scores in a univariate Cox Proportional Hazard model and while the risk score from the traditional radiomics model was not associated with overall survival, the proposed transfer learning-based risk score had significant prognostic value with hazard ratio of 1.86 (95% Confidence Interval: 1.15–3.53, p-value: 0.04).Conclusions: This result suggests that transfer learning-based models may significantly improve prognostic performance in typical small sample size medical imaging studies.https://www.frontiersin.org/article/10.3389/frai.2020.550890/fulltransfer learningradiomicsprognosispancreatic cancersurvival analysis |
| spellingShingle | Yucheng Zhang Edrise M. Lobo-Mueller Paul Karanicolas Steven Gallinger Masoom A. Haider Masoom A. Haider Farzad Khalvati Farzad Khalvati Farzad Khalvati Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma Frontiers in Artificial Intelligence transfer learning radiomics prognosis pancreatic cancer survival analysis |
| title | Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma |
| title_full | Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma |
| title_fullStr | Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma |
| title_full_unstemmed | Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma |
| title_short | Prognostic Value of Transfer Learning Based Features in Resectable Pancreatic Ductal Adenocarcinoma |
| title_sort | prognostic value of transfer learning based features in resectable pancreatic ductal adenocarcinoma |
| topic | transfer learning radiomics prognosis pancreatic cancer survival analysis |
| url | https://www.frontiersin.org/article/10.3389/frai.2020.550890/full |
| work_keys_str_mv | AT yuchengzhang prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT edrisemlobomueller prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT paulkaranicolas prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT stevengallinger prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT masoomahaider prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT masoomahaider prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT farzadkhalvati prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT farzadkhalvati prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma AT farzadkhalvati prognosticvalueoftransferlearningbasedfeaturesinresectablepancreaticductaladenocarcinoma |