Machine Learning-Based Prediction of Glioma IDH Gene Mutation Status
... Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study).
Recent publication
The presence of a certain mutation has an impact on the treatment of gliomas. Scientists at the University Hospital St. Pölten and the Friedrich-Alexander University Erlangen therefore asked themselves how this gene mutation status can be determined precisely before surgery using radiological imaging methods. The researchers trained traditional machine learning algorithms and simple deep learning models with physio-metabolic MRI data and clinical MRI images. The trained AI was used to preoperatively determine gene mutation status in glioma patients at two sites. The two centres use different protocols for the collection of physio-metabolic MRI data. The results show that AI applications can predict the gene mutation status in glioma patients based on physio-metabolic MRI data, but standardisation of measurement techniques is necessary. Thanks to open access funding from KL, the work has been published freely accessible in the journal "Cancers" and was supported by the German Research Foundation (DFG) and KL's "Forschungsimpulse" programme.
Stadlbauer, A, Nikolic, K, Oberndorfer, S, Marhold, F, Kinfe, TM, Meyer-Bäse, A, Bistrian, DA, Schnell, O & Doerfler, A 2024, 'Machine Learning-Based Prediction of Glioma IDH Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study)', Cancers, vol. 16, no. 6. https://doi.org/10.3390/cancers16061102
Prof. Dr. Andreas Stadlbauer
Institute of Diagnostic and Interventional Radiology (University Hospital St. Pölten)
Dr. Katarina Nikolic
Division of Neurology (University Hospital St. Pölten)
Prim. Assoc. Prof. PD Dr. Stefan Oberndorfer FEAN
Division of Neurology (University Hospital St. Pölten)
PD Dr. Franz Marhold
Division of Neurosurgery (University Hospital St. Pölten)
