Comparison of MRI Sequences to Predict ATRX Status Using Radiomics-Based Machine Learning

Abstract

ATRX is an important molecular marker according to the 2021 WHO classification of adult-type diffuse glioma. We aim to predict the ATRX mutation status non-invasively using radiomics-based machine learning models on MRI and to determine which MRI sequence is best suited for this purpose. In this retrospective study, we used MRI images of patients with histologically confirmed glioma, including the sequences T1w without and with the administration of contrast agent, T2w, and the FLAIR. Radiomics features were extracted from the corresponding MRI images by hand-delineated regions of interest. Data partitioning into training data and independent test data was repeated 100 times to avoid random effects. Feature preselection and subsequent model development were performed using Lasso regression. The T2w sequence was found to be the most suitable and the FLAIR sequence the least suitable for predicting ATRX mutations using radiomics-based machine learning models. For the T2w sequence, our seven-feature model developed with Lasso regression achieved a mean AUC of 0.831, a mean accuracy of 0.746, a mean sensitivity of 0.772, and a mean specificity of 0.697. In conclusion, for the prediction of ATRX mutation using radiomics-based machine learning models, the T2w sequence is the most suitable among the commonly used MRI sequences.

Keywords: ATRX; MRI; artificial intelligence; glioma; neuroimaging; radiomics.

Figure 1

Segmentation of a T2w sequence of a Patient with IDH-mutated Oligodendroglioma frontal left, ATRX positive.

Figure 2

Flowchart describing the methodological approach. For each MRI sequence, a total of 15 models are developed with an increasing number (1 to 15) of model features included. Each of these models is developed 100 times, each time with a new data partitioning, and subsequently tested. The final determination of the performance of each model is calculated as the average of the 100 cycles.

Figure 3

Predictability of ATRX mutation status for gliomas using different MRI sequences: Area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for the independent test samples, calculated as means of 100 repetitions (100 cycles) depending on the number of model features included. Feature pre-selection and model construction were performed using Lasso regression.