Trained on ultra-high-resolution annotated MRI datasets, the model segments left and right thalamus and individual nuclei with high accuracy. The framework combines deep learning with anatomical priors to ensure robust and consistent segmentation. Designed to work on standard clinical T1 scans, DeepThalamus outputs volumetric measurements, asymmetry indices, and nucleus-specific data, enabling detailed study of thalamic structure in large neuroimaging datasets.