Abstract:Radiogenomics has shown potential to predict genomic phenotypes from medical images. The development of models using standard-of-care pre-operative MRI images, as opposed to advanced MRI images, enables a broader reach of such models. In this work, a radiogenomics model for IDH mutation status prediction from standard-of-care MRIs in patients with glioma was developed and validated using multicentric data. A cohort of 142 (wild-type: 32.4%) patients with glioma retrieved from the TCIA/TCGA was used to train a logistic regression model to predict the IDH mutation status. The model was evaluated using retrospective data collected in two distinct hospitals, comprising 36 (wild-type: 63.9%) and 53 (wild-type: 75.5%) patients. Model development utilized ROC analysis. Model discrimination and calibration were used for validation. The model yielded an AUC of 0.741 vs. 0.716 vs. 0.938, a sensitivity of 0.784 vs. 0.739 vs. 0.875, and a specificity of 0.657 vs. 0.692 vs. 1.000 on the training, test cohort 1, and test cohort 2, respectively. The assessment of model fairness suggested an unbiased model for age and sex, and calibration tests showed a p < 0.05. These results indicate that the developed model allows the prediction of the IDH mutation status in gliomas using standard-of-care MRI images and does not appear to hold sex and age biases.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to predict the IDH mutation status of glioma patients by developing an end - to - end artificial intelligence system using standard - of - care MRI images. Specifically, the researchers hope to establish a model that can accurately predict the IDH mutation status of patients by extracting and analyzing radiomic features obtained from standard - of - care MRI images. This will not only help clinicians have a more accurate understanding of the patient's condition before surgery, but also guide the formulation of postoperative treatment plans, especially in areas where genomic analysis may not be available. ### Research Background and Objectives - **Background**: Diffuse glioma is the most common primary malignant tumor of the central nervous system in adults, accounting for more than 70% of all primary brain tumors. With the update of diagnostic classifications, especially the revisions of the World Health Organization (WHO) classification of central nervous system tumors in 2016 and 2021, the IDH mutation status has become one of the important criteria for glioma diagnosis. - **Objectives**: To develop a radiomics - based model to predict the IDH mutation status of glioma patients using standard - of - care MRI images, and to verify the effectiveness and fairness of this model through multi - center data. ### Methods - **Data Sources**: The training set used publicly available data from The Cancer Imaging Archive (TCIA) and The Cancer Genome Atlas (TCGA). The external test sets were from two independent European hospitals. - **Processing Flow**: - **Image Pre - processing**: Convert DICOM images to NIfTI format, and perform brain extraction, registration, and segmentation. - **Feature Extraction**: Use the pyradiomics tool to extract a variety of radiomic features from each image sequence and segmented area. - **Model Training**: Use the five - fold stratified cross - validation method to train a logistic regression classifier and optimize the ridge regularization intensity to reduce overfitting. - **Evaluation**: The model was evaluated on two external test sets, and the main indicators include AUC, sensitivity, specificity, and accuracy. ### Results - **Model Performance**: - **Training Set**: AUC is 0.741, sensitivity is 0.784, and specificity is 0.657. - **Test Set 1**: AUC is 0.716, sensitivity is 0.739, and specificity is 0.692. - **Test Set 2**: AUC is 0.938, sensitivity is 0.875, and specificity is 1.000. - **Fairness Evaluation**: The model has no obvious bias in terms of age and gender. ### Discussion - **Significance**: The model developed in this study can accurately predict the IDH mutation status of glioma patients based on standard - of - care MRI images, which has important clinical application value. Especially in cases where genomic analysis is not available, this model can provide valuable diagnostic information. - **Limitations**: Although the model performs well on the external test sets, the calibration test results show that the model needs further calibration when used in external institutions. In addition, the small sample size may affect the universality of the results. ### Conclusion This study successfully developed and verified a radiomics - based model for predicting the IDH mutation status of glioma patients. The model performs well on multi - center data, has no obvious bias in terms of age and gender, and has high clinical application potential.

Development of End-to-End AI–Based MRI Image Analysis System for Predicting IDH Mutation Status of Patients with Gliomas: Multicentric Validation

Using Radiomics Based on Multicenter Magnetic Resonance Images to Predict Isocitrate Dehydrogenase Mutation Status of Gliomas.

MRI Radiomic Features to Predict IDH1 Mutation Status in Gliomas: A Machine Learning Approach using Gradient Tree Boosting

Multiregional Radiomics Profiling from Multiparametric MRI: Identifying an Imaging Predictor of IDH1 Mutation Status in Glioblastoma

Machine Learning-Based Prediction of Glioma IDH Gene Mutation Status Using Physio-Metabolic MRI of Oxygen Metabolism and Neovascularization (A Bicenter Study)

Prediction of IDH Mutation Status of Glioma Based on Multimodal MRI Images

Comparison of MRI Sequences to Predict IDH Mutation Status in Gliomas Using Radiomics-Based Machine Learning

Qualitative and Quantitative MRI Analysis in IDH1 Genotype Prediction of Lower-Grade Gliomas: A Machine Learning Approach.

Predicting IDH Mutation Status in Low-Grade Gliomas Based on Optimal Radiomic Features Combined with Multi-Sequence Magnetic Resonance Imaging

Machine learning reveals multimodal MRI patterns predictive of isocitrate dehydrogenase and 1p/19q status in diffuse low- and high-grade gliomas

Prediction of the Molecular Subtype of IDH Mutation Combined with MGMT Promoter Methylation in Gliomas via Radiomics Based on Preoperative MRI

Easy-to-use machine learning system for the prediction of IDH mutation and 1p/19q codeletion using MRI images of adult-type diffuse gliomas

Prediction of IDH1 Gene Mutation by a Nomogram Based on Multiparametric and Multiregional MR Images

Predicting Isocitrate Dehydrogenase ( IDH) Mutation Status in Gliomas Using Multiparameter MRI Radiomics Features

Predictive machine learning models based on VASARI features for WHO grading, isocitrate dehydrogenase mutation, and 1p19q co-deletion status: a multicenter study

Completely non-invasive prediction of IDH mutation status based on preoperative native CT images

Predicting Isocitrate Dehydrogenase Genotype, Histological Phenotype, and Ki-67 Expression Level in Diffuse Gliomas with an Advanced Contrast Analysis of Magnetic Resonance Imaging Sequences.

Diagnostic Performance of Deep Learning for Predicting Gliomas' IDH and 1p/19q Status in MRI: A Systematic Review and Meta-Analysis

Combination of MRI-based prediction and CRISPR/Cas12a-based detection for IDH genotyping in glioma

Deep Learning Can Differentiate IDH-Mutant from IDH-Wild GBM

Predictive Role of the Apparent Diffusion Coefficient and MRI Morphologic Features on IDH Status in Patients With Diffuse Glioma: A Retrospective Cross-Sectional Study