Current browse context:
q-bio.QM
Change to browse by:
References & Citations
Quantitative Biology > Quantitative Methods
Title: Tractometry-based Anomaly Detection for Single-subject White Matter Analysis
(Submitted on 22 May 2020 (v1), last revised 25 May 2020 (this version, v2))
Abstract: There is an urgent need for a paradigm shift from group-wise comparisons to individual diagnosis in diffusion MRI (dMRI) to enable the analysis of rare cases and clinically-heterogeneous groups. Deep autoencoders have shown great potential to detect anomalies in neuroimaging data. We present a framework that operates on the manifold of white matter (WM) pathways to learn normative microstructural features, and discriminate those at genetic risk from controls in a paediatric population.
Submission history
From: Maxime Chamberland [view email][v1] Fri, 22 May 2020 09:50:22 GMT (2516kb,D)
[v2] Mon, 25 May 2020 18:59:34 GMT (2516kb,D)
Link back to: arXiv, form interface, contact.