Classifyber, a robust streamline-based linear classifier for white matter bundle segmentation

• DOI: 10.1101/2020.02.10.942714
• Publication Date: 2020-02-12T21:35:13Z
• Authors: Giulia Bertò; Daniel Bullock; Pietro Astolfi; Soichi Hayashi; Luca Zigiotto; Luciano Annicchia...; Francesco Corsini; Alessandro De Ben...; Silvio Sarubbo; Franco Pestilli; Paolo Avesani; Emanuele Olivetti
• Keywords: Connectomics; Human Connectome Project

Abstract Virtual delineation of white matter bundles in the human brain is paramount importance for multiple applications, such as pre-surgical planning and connectomics. A substantial body literature related to methods that automatically segment from diffusion Magnetic Resonance Imaging (dMRI) data indirectly, by exploiting either idea connectivity between regions or geometry fiber paths obtained with tractography techniques, or, directly, through information volumetric data. Despite remarkable improvement automatic segmentation over years, their quality not yet satisfactory, especially when dealing datasets very diverse characteristics, different tracking methods, bundle sizes quality. In this work, we propose a novel, supervised streamline-based method, called Classifyber, which combines atlases, patterns, into simple linear model. With wide range experiments on span research clinical domains, show Classifyber substantially improves compared other state-of-the-art and, more importantly, it robust across settings. We provide an implementation proposed method open source code, well web service.