Deep learning segmentation of the choroid plexus from structural magnetic resonance imaging (MRI): validation and normative ranges across the adult lifespan

Jarrod J. Eisma,Colin D. McKnight,Kilian Hett,Jason Elenberger,Caleb J. Han,Alexander K. Song,Ciaran Considine,Daniel O. Claassen,Manus J. Donahue
DOI: https://doi.org/10.1186/s12987-024-00525-9
2024-03-02
Fluids and Barriers of the CNS
Abstract:The choroid plexus functions as the blood-cerebrospinal fluid (CSF) barrier, plays an important role in CSF production and circulation, and has gained increased attention in light of the recent elucidation of CSF circulation dysfunction in neurodegenerative conditions. However, methods for routinely quantifying choroid plexus volume are suboptimal and require technical improvements and validation. Here, we propose three deep learning models that can segment the choroid plexus from commonly-acquired anatomical MRI data and report performance metrics and changes across the adult lifespan.
neurosciences
What problem does this paper attempt to address?