Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study)
The mutational status of the isocitrate dehydrogenase (<i>IDH</i>) gene plays a key role in the treatment of glioma patients because it is known to affect energy metabolism pathways relevant to glioma. Physio-metabolic magnetic resonance imaging (MRI) enables the non-invasive analysis of...
Main Authors: | , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2024-03-01
|
Series: | Cancers |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-6694/16/6/1102 |
_version_ | 1797241745464885248 |
---|---|
author | Andreas Stadlbauer Katarina Nikolic Stefan Oberndorfer Franz Marhold Thomas M. Kinfe Anke Meyer-Bäse Diana Alina Bistrian Oliver Schnell Arnd Doerfler |
author_facet | Andreas Stadlbauer Katarina Nikolic Stefan Oberndorfer Franz Marhold Thomas M. Kinfe Anke Meyer-Bäse Diana Alina Bistrian Oliver Schnell Arnd Doerfler |
author_sort | Andreas Stadlbauer |
collection | DOAJ |
description | The mutational status of the isocitrate dehydrogenase (<i>IDH</i>) gene plays a key role in the treatment of glioma patients because it is known to affect energy metabolism pathways relevant to glioma. Physio-metabolic magnetic resonance imaging (MRI) enables the non-invasive analysis of oxygen metabolism and tissue hypoxia as well as associated neovascularization and microvascular architecture. However, evaluating such complex neuroimaging data requires computational support. Traditional machine learning algorithms and simple deep learning models were trained with radiomic features from clinical MRI (cMRI) or physio-metabolic MRI data. A total of 215 patients (first center: 166 participants + 16 participants for independent internal testing of the algorithms versus second site: 33 participants for independent external testing) were enrolled using two different physio-metabolic MRI protocols. The algorithms trained with physio-metabolic data demonstrated the best classification performance in independent internal testing: precision, 91.7%; accuracy, 87.5%; area under the receiver operating curve (AUROC), 0.979. In external testing, traditional machine learning models trained with cMRI data exhibited the best <i>IDH</i> classification results: precision, 84.9%; accuracy, 81.8%; and AUROC, 0.879. The poor performance for the physio-metabolic MRI approach appears to be explainable by site-dependent differences in data acquisition methodologies. The physio-metabolic MRI approach potentially supports reliable classification of <i>IDH</i> gene status in the presurgical stage of glioma patients. However, non-standardized protocols limit the level of evidence and underlie the need for a reproducible framework of data acquisition techniques. |
first_indexed | 2024-04-24T18:28:13Z |
format | Article |
id | doaj.art-a7721e0ffe2046bfab784c19bb8af1d3 |
institution | Directory Open Access Journal |
issn | 2072-6694 |
language | English |
last_indexed | 2024-04-24T18:28:13Z |
publishDate | 2024-03-01 |
publisher | MDPI AG |
record_format | Article |
series | Cancers |
spelling | doaj.art-a7721e0ffe2046bfab784c19bb8af1d32024-03-27T13:29:50ZengMDPI AGCancers2072-66942024-03-01166110210.3390/cancers16061102Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study)Andreas Stadlbauer0Katarina Nikolic1Stefan Oberndorfer2Franz Marhold3Thomas M. Kinfe4Anke Meyer-Bäse5Diana Alina Bistrian6Oliver Schnell7Arnd Doerfler8Karl Landsteiner University of Health Sciences, 3500 Krems, AustriaKarl Landsteiner University of Health Sciences, 3500 Krems, AustriaKarl Landsteiner University of Health Sciences, 3500 Krems, AustriaKarl Landsteiner University of Health Sciences, 3500 Krems, AustriaDepartment of Neurosurgery, Universitätsklinikum Erlangen, Friedrich-Alexander University (FAU) Erlangen-Nürnberg, 91054 Erlangen, GermanyDepartment of Scientific Computing, Florida State University, 400 Dirac Science Library Tallahassee, Tallahassee, FL 32306-4120, USADepartment of Electrical Engineering and Industrial Informatics, Politehnica University of Timisoara, 300006 Timișoara, RomaniaDepartment of Neurosurgery, Universitätsklinikum Erlangen, Friedrich-Alexander University (FAU) Erlangen-Nürnberg, 91054 Erlangen, GermanyDepartment of Neuroradiology, Universitätsklinikum Erlangen, Friedrich-Alexander University (FAU) Erlangen-Nürnberg, 91054 Erlangen, GermanyThe mutational status of the isocitrate dehydrogenase (<i>IDH</i>) gene plays a key role in the treatment of glioma patients because it is known to affect energy metabolism pathways relevant to glioma. Physio-metabolic magnetic resonance imaging (MRI) enables the non-invasive analysis of oxygen metabolism and tissue hypoxia as well as associated neovascularization and microvascular architecture. However, evaluating such complex neuroimaging data requires computational support. Traditional machine learning algorithms and simple deep learning models were trained with radiomic features from clinical MRI (cMRI) or physio-metabolic MRI data. A total of 215 patients (first center: 166 participants + 16 participants for independent internal testing of the algorithms versus second site: 33 participants for independent external testing) were enrolled using two different physio-metabolic MRI protocols. The algorithms trained with physio-metabolic data demonstrated the best classification performance in independent internal testing: precision, 91.7%; accuracy, 87.5%; area under the receiver operating curve (AUROC), 0.979. In external testing, traditional machine learning models trained with cMRI data exhibited the best <i>IDH</i> classification results: precision, 84.9%; accuracy, 81.8%; and AUROC, 0.879. The poor performance for the physio-metabolic MRI approach appears to be explainable by site-dependent differences in data acquisition methodologies. The physio-metabolic MRI approach potentially supports reliable classification of <i>IDH</i> gene status in the presurgical stage of glioma patients. However, non-standardized protocols limit the level of evidence and underlie the need for a reproducible framework of data acquisition techniques.https://www.mdpi.com/2072-6694/16/6/1102artificial intelligencedeep learninggliomaisocitrate dehydrogenase gene<i>IDH</i> gene mutationneurooncology |
spellingShingle | Andreas Stadlbauer Katarina Nikolic Stefan Oberndorfer Franz Marhold Thomas M. Kinfe Anke Meyer-Bäse Diana Alina Bistrian Oliver Schnell Arnd Doerfler Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study) Cancers artificial intelligence deep learning glioma isocitrate dehydrogenase gene <i>IDH</i> gene mutation neurooncology |
title | Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study) |
title_full | Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study) |
title_fullStr | Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study) |
title_full_unstemmed | Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study) |
title_short | Machine Learning-Based Prediction of Glioma <i>IDH</i> Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study) |
title_sort | machine learning based prediction of glioma i idh i gene mutation status using physio metabolic mri of oxygen metabolism and neovascularization a bicenter study |
topic | artificial intelligence deep learning glioma isocitrate dehydrogenase gene <i>IDH</i> gene mutation neurooncology |
url | https://www.mdpi.com/2072-6694/16/6/1102 |
work_keys_str_mv | AT andreasstadlbauer machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT katarinanikolic machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT stefanoberndorfer machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT franzmarhold machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT thomasmkinfe machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT ankemeyerbase machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT dianaalinabistrian machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT oliverschnell machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy AT arnddoerfler machinelearningbasedpredictionofgliomaiidhigenemutationstatususingphysiometabolicmriofoxygenmetabolismandneovascularizationabicenterstudy |