End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT
(1) Background: CT perfusion (CTP) is used to quantify cerebral hypoperfusion in acute ischemic stroke. Conventional attenuation curve analysis is not standardized and might require input from expert users, hampering clinical application. This study aims to bypass conventional tracer-kinetic analysi...
Main Authors: | , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-05-01
|
Series: | Diagnostics |
Subjects: | |
Online Access: | https://www.mdpi.com/2075-4418/12/5/1142 |
_version_ | 1797500471655530496 |
---|---|
author | Andreas Mittermeier Paul Reidler Matthias P. Fabritius Balthasar Schachtner Philipp Wesp Birgit Ertl-Wagner Olaf Dietrich Jens Ricke Lars Kellert Steffen Tiedt Wolfgang G. Kunz Michael Ingrisch |
author_facet | Andreas Mittermeier Paul Reidler Matthias P. Fabritius Balthasar Schachtner Philipp Wesp Birgit Ertl-Wagner Olaf Dietrich Jens Ricke Lars Kellert Steffen Tiedt Wolfgang G. Kunz Michael Ingrisch |
author_sort | Andreas Mittermeier |
collection | DOAJ |
description | (1) Background: CT perfusion (CTP) is used to quantify cerebral hypoperfusion in acute ischemic stroke. Conventional attenuation curve analysis is not standardized and might require input from expert users, hampering clinical application. This study aims to bypass conventional tracer-kinetic analysis with an end-to-end deep learning model to directly categorize patients by stroke core volume from raw, slice-reduced CTP data. (2) Methods: In this retrospective analysis, we included patients with acute ischemic stroke due to proximal occlusion of the anterior circulation who underwent CTP imaging. A novel convolutional neural network was implemented to extract spatial and temporal features from time-resolved imaging data. In a classification task, the network categorized patients into small or large core. In ten-fold cross-validation, the network was repeatedly trained, evaluated, and tested, using the area under the receiver operating characteristic curve (ROC-AUC). A final model was created in an ensemble approach and independently validated on an external dataset. (3) Results: 217 patients were included in the training cohort and 23 patients in the independent test cohort. Median core volume was 32.4 mL and was used as threshold value for the binary classification task. Model performance yielded a mean (SD) ROC-AUC of 0.72 (0.10) for the test folds. External independent validation resulted in an ensembled mean ROC-AUC of 0.61. (4) Conclusions: In this proof-of-concept study, the proposed end-to-end deep learning approach bypasses conventional perfusion analysis and allows to predict dichotomized infarction core volume solely from slice-reduced CTP images without underlying tracer kinetic assumptions. Further studies can easily extend to additional clinically relevant endpoints. |
first_indexed | 2024-03-10T03:03:21Z |
format | Article |
id | doaj.art-e4889f92bf894574a2a89335e7cf55a3 |
institution | Directory Open Access Journal |
issn | 2075-4418 |
language | English |
last_indexed | 2024-03-10T03:03:21Z |
publishDate | 2022-05-01 |
publisher | MDPI AG |
record_format | Article |
series | Diagnostics |
spelling | doaj.art-e4889f92bf894574a2a89335e7cf55a32023-11-23T10:39:58ZengMDPI AGDiagnostics2075-44182022-05-01125114210.3390/diagnostics12051142End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CTAndreas Mittermeier0Paul Reidler1Matthias P. Fabritius2Balthasar Schachtner3Philipp Wesp4Birgit Ertl-Wagner5Olaf Dietrich6Jens Ricke7Lars Kellert8Steffen Tiedt9Wolfgang G. Kunz10Michael Ingrisch11Department of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Diagnostic Imaging, The Hospital for Sick Children, University of Toronto, Toronto, ON M5G 1X8, CanadaDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Neurology, University Hospital, LMU Munich, 81377 Munich, GermanyInstitute for Stroke and Dementia Research, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, GermanyDepartment of Radiology, University Hospital, LMU Munich, 81377 Munich, Germany(1) Background: CT perfusion (CTP) is used to quantify cerebral hypoperfusion in acute ischemic stroke. Conventional attenuation curve analysis is not standardized and might require input from expert users, hampering clinical application. This study aims to bypass conventional tracer-kinetic analysis with an end-to-end deep learning model to directly categorize patients by stroke core volume from raw, slice-reduced CTP data. (2) Methods: In this retrospective analysis, we included patients with acute ischemic stroke due to proximal occlusion of the anterior circulation who underwent CTP imaging. A novel convolutional neural network was implemented to extract spatial and temporal features from time-resolved imaging data. In a classification task, the network categorized patients into small or large core. In ten-fold cross-validation, the network was repeatedly trained, evaluated, and tested, using the area under the receiver operating characteristic curve (ROC-AUC). A final model was created in an ensemble approach and independently validated on an external dataset. (3) Results: 217 patients were included in the training cohort and 23 patients in the independent test cohort. Median core volume was 32.4 mL and was used as threshold value for the binary classification task. Model performance yielded a mean (SD) ROC-AUC of 0.72 (0.10) for the test folds. External independent validation resulted in an ensembled mean ROC-AUC of 0.61. (4) Conclusions: In this proof-of-concept study, the proposed end-to-end deep learning approach bypasses conventional perfusion analysis and allows to predict dichotomized infarction core volume solely from slice-reduced CTP images without underlying tracer kinetic assumptions. Further studies can easily extend to additional clinically relevant endpoints.https://www.mdpi.com/2075-4418/12/5/1142CT perfusionstrokedeep learningcontrast-enhanced perfusion imagingconvolutional neural networksend-to-end modeling |
spellingShingle | Andreas Mittermeier Paul Reidler Matthias P. Fabritius Balthasar Schachtner Philipp Wesp Birgit Ertl-Wagner Olaf Dietrich Jens Ricke Lars Kellert Steffen Tiedt Wolfgang G. Kunz Michael Ingrisch End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT Diagnostics CT perfusion stroke deep learning contrast-enhanced perfusion imaging convolutional neural networks end-to-end modeling |
title | End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT |
title_full | End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT |
title_fullStr | End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT |
title_full_unstemmed | End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT |
title_short | End-to-End Deep Learning Approach for Perfusion Data: A Proof-of-Concept Study to Classify Core Volume in Stroke CT |
title_sort | end to end deep learning approach for perfusion data a proof of concept study to classify core volume in stroke ct |
topic | CT perfusion stroke deep learning contrast-enhanced perfusion imaging convolutional neural networks end-to-end modeling |
url | https://www.mdpi.com/2075-4418/12/5/1142 |
work_keys_str_mv | AT andreasmittermeier endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT paulreidler endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT matthiaspfabritius endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT balthasarschachtner endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT philippwesp endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT birgitertlwagner endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT olafdietrich endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT jensricke endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT larskellert endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT steffentiedt endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT wolfganggkunz endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect AT michaelingrisch endtoenddeeplearningapproachforperfusiondataaproofofconceptstudytoclassifycorevolumeinstrokect |