Radiomic tractometry: a rich and tract-specific class of imaging biomarkers for neuroscience and medical applications

Tract-specific microstructural analysis of the brain white matter using diffusion MRI is a driver for neuroscientific discovery with a wide range of applications. Current analysis pipelines have conceptual limitations that narrow their applicability and hamper subject-level analysis and predictions. Radiomic tractometry (RadTract) improves upon that, enabling the extraction and analysis of comprehensive and highly informative microstructural feature sets where previous approaches were restricted to bare summary statistics. We demonstrate the added value in a series of neuroscientific applications, including diagnostic tasks as well as the prediction of demographic and clinical measures across various datasets. Being published as an open and easy-to-use Python package, RadTract could spark the establishment of a new generation of tract-specific imaging biomarkers, with direct benefits for a range of applications from basic neuroscience to medical research.