Prediction of gender from structural MRI images using Multiscale ShuffleNet Extreme Learning Machine

In recent years, the application of machine learning techniques to medical imaging data has shown promising results in various clinical tasks. One such task is the prediction of gender from structural MRI (magnetic resonance imaging) images, which holds potential implications for personalized medicine and understanding neurobiological differences between genders. In this study, we propose a novel approach utilizing Multiscale ShuffleNet Extreme Learning Machine (MSEL), a fusion of Multiscale feature extraction and ShuffleNet architecture with Extreme Learning Machine classifier. We demonstrate the effectiveness of our method on a large dataset of structural MRI images, employing state-of-the-art preprocessing techniques and feature extraction methods. Our results indicate significant accuracy and robustness in gender prediction, outperforming existing methodologies. Furthermore, we conduct comprehensive analyses to investigate the contribution of different components in our proposed framework, shedding light on the underlying mechanisms of gender-related brain structural differences. Overall, our study presents a promising avenue for utilizing advanced machine learning techniques in neuroimaging research, with potential applications in clinical diagnostics and personalized healthcare.