A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging
Abstract Background To develop a supervised machine learning (ML) algorithm predicting above- versus below-median overall survival (OS) from diffusion-weighted imaging-derived radiomic features in patients with pancreatic ductal adenocarcinoma (PDAC). Methods One hundred two patients with histopatho...
Main Authors: | , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
SpringerOpen
2019-10-01
|
Series: | European Radiology Experimental |
Subjects: | |
Online Access: | http://link.springer.com/article/10.1186/s41747-019-0119-0 |
id |
doaj-d7a868f772a24161b5c328977cae37ee |
---|---|
record_format |
Article |
spelling |
doaj-d7a868f772a24161b5c328977cae37ee2020-11-25T03:54:04ZengSpringerOpenEuropean Radiology Experimental2509-92802019-10-01311910.1186/s41747-019-0119-0A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imagingGeorgios Kaissis0Sebastian Ziegelmayer1Fabian Lohöfer2Hana Algül3Matthias Eiber4Wilko Weichert5Roland Schmid6Helmut Friess7Ernst Rummeny8Donna Ankerst9Jens Siveke10Rickmer Braren11Department of Diagnostic and Interventional Radiology, Faculty of Medicine, Klinikum rechts der Isar, Technical University of MunichDepartment of Diagnostic and Interventional Radiology, Faculty of Medicine, Klinikum rechts der Isar, Technical University of MunichDepartment of Diagnostic and Interventional Radiology, Faculty of Medicine, Klinikum rechts der Isar, Technical University of MunichDepartment of Internal Medicine II, Faculty of Medicine, Technical University of MunichDepartment of Nuclear Medicine, Faculty of Medicine, Technical University of MunichDepartment of Pathology, Faculty of Medicine, Technical University of MunichDepartment of Internal Medicine II, Faculty of Medicine, Technical University of MunichDepartment of Surgery, Faculty of Medicine, Technical University of MunichDepartment of Diagnostic and Interventional Radiology, Faculty of Medicine, Klinikum rechts der Isar, Technical University of MunichDepartment of Mathematics, Technical University of MunichWest German Cancer Center, University of EssenDepartment of Diagnostic and Interventional Radiology, Faculty of Medicine, Klinikum rechts der Isar, Technical University of MunichAbstract Background To develop a supervised machine learning (ML) algorithm predicting above- versus below-median overall survival (OS) from diffusion-weighted imaging-derived radiomic features in patients with pancreatic ductal adenocarcinoma (PDAC). Methods One hundred two patients with histopathologically proven PDAC were retrospectively assessed as training cohort, and 30 prospectively accrued and retrospectively enrolled patients served as independent validation cohort (IVC). Tumors were segmented on preoperative apparent diffusion coefficient (ADC) maps, and radiomic features were extracted. A random forest ML algorithm was fit to the training cohort and tested in the IVC. Histopathological subtype of tumor samples was assessed by immunohistochemistry in 21 IVC patients. Individual radiomic feature importance was evaluated by assessment of tree node Gini impurity decrease and recursive feature elimination. Fisher’s exact test, 95% confidence intervals (CI), and receiver operating characteristic area under the curve (ROC-AUC) were used. Results The ML algorithm achieved 87% sensitivity (95% IC 67.3–92.7), 80% specificity (95% CI 74.0–86.7), and ROC-AUC 90% for the prediction of above- versus below-median OS in the IVC. Heterogeneity-related features were highly ranked by the model. Of the 21 patients with determined histopathological subtype, 8/9 patients predicted to experience below-median OS exhibited the quasi-mesenchymal subtype, whilst 11/12 patients predicted to experience above-median OS exhibited a non-quasi-mesenchymal subtype (p < 0.001). Conclusion ML application to ADC radiomics allowed OS prediction with a high diagnostic accuracy in an IVC. The high overlap of clinically relevant histopathological subtypes with model predictions underlines the potential of quantitative imaging in PDAC pre-operative subtyping and prognosis.http://link.springer.com/article/10.1186/s41747-019-0119-0Machine learningDiffusion magnetic resonance imagingPancreatic carcinomaRadiomicsSurvival analysis |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Georgios Kaissis Sebastian Ziegelmayer Fabian Lohöfer Hana Algül Matthias Eiber Wilko Weichert Roland Schmid Helmut Friess Ernst Rummeny Donna Ankerst Jens Siveke Rickmer Braren |
spellingShingle |
Georgios Kaissis Sebastian Ziegelmayer Fabian Lohöfer Hana Algül Matthias Eiber Wilko Weichert Roland Schmid Helmut Friess Ernst Rummeny Donna Ankerst Jens Siveke Rickmer Braren A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging European Radiology Experimental Machine learning Diffusion magnetic resonance imaging Pancreatic carcinoma Radiomics Survival analysis |
author_facet |
Georgios Kaissis Sebastian Ziegelmayer Fabian Lohöfer Hana Algül Matthias Eiber Wilko Weichert Roland Schmid Helmut Friess Ernst Rummeny Donna Ankerst Jens Siveke Rickmer Braren |
author_sort |
Georgios Kaissis |
title |
A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging |
title_short |
A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging |
title_full |
A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging |
title_fullStr |
A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging |
title_full_unstemmed |
A machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging |
title_sort |
machine learning model for the prediction of survival and tumor subtype in pancreatic ductal adenocarcinoma from preoperative diffusion-weighted imaging |
publisher |
SpringerOpen |
series |
European Radiology Experimental |
issn |
2509-9280 |
publishDate |
2019-10-01 |
description |
Abstract Background To develop a supervised machine learning (ML) algorithm predicting above- versus below-median overall survival (OS) from diffusion-weighted imaging-derived radiomic features in patients with pancreatic ductal adenocarcinoma (PDAC). Methods One hundred two patients with histopathologically proven PDAC were retrospectively assessed as training cohort, and 30 prospectively accrued and retrospectively enrolled patients served as independent validation cohort (IVC). Tumors were segmented on preoperative apparent diffusion coefficient (ADC) maps, and radiomic features were extracted. A random forest ML algorithm was fit to the training cohort and tested in the IVC. Histopathological subtype of tumor samples was assessed by immunohistochemistry in 21 IVC patients. Individual radiomic feature importance was evaluated by assessment of tree node Gini impurity decrease and recursive feature elimination. Fisher’s exact test, 95% confidence intervals (CI), and receiver operating characteristic area under the curve (ROC-AUC) were used. Results The ML algorithm achieved 87% sensitivity (95% IC 67.3–92.7), 80% specificity (95% CI 74.0–86.7), and ROC-AUC 90% for the prediction of above- versus below-median OS in the IVC. Heterogeneity-related features were highly ranked by the model. Of the 21 patients with determined histopathological subtype, 8/9 patients predicted to experience below-median OS exhibited the quasi-mesenchymal subtype, whilst 11/12 patients predicted to experience above-median OS exhibited a non-quasi-mesenchymal subtype (p < 0.001). Conclusion ML application to ADC radiomics allowed OS prediction with a high diagnostic accuracy in an IVC. The high overlap of clinically relevant histopathological subtypes with model predictions underlines the potential of quantitative imaging in PDAC pre-operative subtyping and prognosis. |
topic |
Machine learning Diffusion magnetic resonance imaging Pancreatic carcinoma Radiomics Survival analysis |
url |
http://link.springer.com/article/10.1186/s41747-019-0119-0 |
work_keys_str_mv |
AT georgioskaissis amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT sebastianziegelmayer amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT fabianlohofer amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT hanaalgul amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT matthiaseiber amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT wilkoweichert amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT rolandschmid amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT helmutfriess amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT ernstrummeny amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT donnaankerst amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT jenssiveke amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT rickmerbraren amachinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT georgioskaissis machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT sebastianziegelmayer machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT fabianlohofer machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT hanaalgul machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT matthiaseiber machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT wilkoweichert machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT rolandschmid machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT helmutfriess machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT ernstrummeny machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT donnaankerst machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT jenssiveke machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging AT rickmerbraren machinelearningmodelforthepredictionofsurvivalandtumorsubtypeinpancreaticductaladenocarcinomafrompreoperativediffusionweightedimaging |
_version_ |
1724474961420091392 |