Assessing the impact of subject-specific masks on reliability of subcortical connectivity

Poster Session C - Sunday, April 14, 2024, 5:00 – 7:00 pm EDT, Sheraton Hall ABC

Alexandra Fischbach¹ (fischbach.a@northeastern.edu), Hallee Shearer¹, Ajay Satpute¹, Karen Quigley¹, Jordan Theriault^1,3, Lisa Feldman Barrett^1,2,3, Stephanie Noble¹; ¹Northeastern University, ²Massachusetts General Hospital and Harvard Medical School, ³Athinoula A. Martinos Center for Biomedical Imaging

The connection between the periaqueductal gray (PAG) and the amygdala is regarded as a key pathway for regulating fear processing and pain modulation. However, due to the PAG’s location within the subcortex surrounded by cerebrospinal fluid, the PAG is highly susceptible to physiological distortions and noise. Moreover, due to the PAG’s small size and inter-individual variability, it is unclear how effectively standard group-derived masks can isolate PAG signal, particularly at standard field strengths. To address this gap, we examined how the use of subject-specific PAG masks influence the reliability of 7-Tesla PAG-amygdala functional connectivity across 3 consecutive resting-state runs in humans (N=48) using the intraclass correlation coefficient (ICC). Subject- and run-specific PAG masks were created by identifying the cerebral aqueduct as voxels with high variance at the individual level, dilating this selection by 2mm, and removing the aqueduct. Contrary to our predictions, we found comparable levels of reliability of PAG-amygdala connectivity when using subject-specific (ICC = 0.532) and group (ICC = 0.513) PAG masks. Although this suggests a minimal impact of masking strategy on reliability, given prior research advocating for subject-specific masks in other contexts, future investigations are needed to comprehensively compare the benefits of functionally defining the PAG. Furthermore, the relatively good reliability found here suggests an exception to the low reliability often reported for subcortical connectivity, warranting further investigation. Evaluating which factors influence PAG reliability is critical for delineating the role of the PAG and related structures in ongoing investigations of emotional and physical disorders.

Topic Area: METHODS: Neuroimaging