Machine learning for detecting age and sex effects on corpus callosum volume

Aim This study aims to investigate whether machine learning (ML) could identify the volumetric changes in the corpus callosum (CC) and its sub-regions (genu, body, and splenium) due to aging and sex differences. Material and Methods Brain magnetic resonance imaging (MRI) images obtained from 301 healthy male and female subjects were used in the study. The volume measurements of the corpus callosum and its sub-regions were calculated from MRI images using MRICloud software. The classifications of age (young/adult) and sex (female/male) were performed using the Classification Learner Tool in MATLAB 2020b (MathWorks, Natick, MA, USA). Classifiers including k-Nearest Neighbor (KNN), support vector machine (SVM), decision tree, naïve Bayes, logistic regression, and Ensemble classifiers were evaluated using 10-fold cross-validation. Results Machine learning classification revealed moderate accuracy in distinguishing sex (best: Fine Gaussian SVM, 65.4% accuracy, AUC = 0.60) and higher accuracy in distinguishing age groups (young/adult; best: Fine Gaussian SVM, 83.7% accuracy, area under the curve (AUC) = 0.67) based on corpus callosum and its sub-region volumes derived from MRI. Conclusion This preliminary data suggests that ML can provide indications to assess the impact of age and sex on corpus callosum volume based on MRI data. The observed classification accuracies, particularly for sex, suggest that larger datasets are needed to enhance the accuracy of these models in future investigations.