A Reproducible Pipeline for Parcellation of the Anterior Limb of the Internal Capsule
Karianne Sretavan1, Henry Braun2, Zoe Liu3
1Graduate Program in Neuroscience, University of Minnesota, Minneapolis, Minnesota; Center for Magnetic Resonance Research, Department of Radiology, University of Minnesota, Minneapolis, Minnesota.
Background:
The anterior limb of the internal capsule (ALIC) is a white matter structure that connects the prefrontal cortex (PFC) to the brainstem, thalamus, and subthalamic nucleus. It is a target for deep brain stimulation for obsessive-compulsive disorder. There is strong interest in improving deep brain stimulation targeting by using diffusion tractography to reconstruct and target specific ALIC fiber pathways, but this methodology is susceptible to errors and lacks validation. To address these limitations, we developed a novel diffusion tractography pipeline that generates reliable and biologically validated ALIC white matter reconstructions.
Methods:
Following algorithm development and refinement, we analyzed 43 control participants, each with 2 sets of 3T magnetic resonance imaging data and a subset of 5 control participants with 7T data from the Human Connectome Project. We generated 22 segmented ALIC fiber bundles (11 per hemisphere) based on PFC regions of interest, and we analyzed the relationships among bundles.
Results:
We successfully reproduced the topographies established by previous anatomical work using images acquired at both 3T and 7T. Quantitative assessment demonstrated significantly smaller intraparticipant variability than interparticipant variability for both test and retest groups across all but one PFC region. We examined the overlap between fibers from different PFC regions and a response tract for obsessive-compulsive disorder deep brain stimulation, and we reconstructed the PFC hyperdirect pathway using a modified version of our pipeline.
Conclusions:
Our diffusion magnetic resonance imaging algorithm reliably generates biologically validated ALIC white matter reconstructions, thereby allowing for more precise modeling of fibers for neuromodulation therapies.
