| Abstrakt: |
Pain is a multifaceted experience incorporating volumetric changes in sensorimotor and pain-processing brain networks. However, divergent findings of region and direction of volume change contribute to the ambiguity surrounding differences in gray matter (GM) volume between individuals with chronic low back pain (cLBP) and healthy controls (HC). This prospective observational study aimed to identify a distinctive pattern of brain regions that could accurately classify cLBP and HC based on the GM volume. Fifty-four cLBP (age=31.6±11.5, F=35) and 50 age- and sex-matched HC (30.2±11.2, F=34) underwent brain MRI (T1-weighted MPRAGE scan;1mm3) on a 3T MRI. T1-Weighted images were processed with steps including brain extraction, tissue segmentation, tissue extraction, and normalization to obtain regional GM volumes. As expected, whole brain GM volume was not significantly different (p=.824) between groups. A stepwise discriminant analysis of 20 bilateral regions of interest (ROI) representing sensorimotor and pain-processing networks was conducted. The discriminant analysis identified a subset of ROIs that separated the groups (accuracy=80%, sensitivity=78%, specificity=82%, cross-validation accuracy=67.3%). The model was comprised of sensorimotor (Bilateral Precuneus, R M1, and L M1-medial), nociceptive (R S1-medial), and context-emotional (bilateral Amygdala, R Nucleus accumbens, L Pallidum, R Anterior Insular cortex, L Ventromedial Prefrontal cortex) pain processing regions suggesting an interaction between the movement and pain systems. The structure matrix indicated R S1-medial (-.363) and R Amygdala (.337) were the main unique contributors to the model separating the groups. Future research should aim to validate this predictive model across different cLBP populations. Funding: R01HD095959. |
