A subject-specific framework for in vivo myeloarchitectonic analysis using high resolution quantitative MRI
The concentration of transmitter receptors varies between different locations in the human cerebral cortex, but also between the different cortical layers within the same area. Analyzing the regional differences in the laminar distribution patterns of various neurotransmitter receptor binding sites by means of quantitative receptor autoradiography may thus provide a functionally relevant insight into the organization of the cortex. Here we introduce an approach to the analysis of in vitro receptor autoradiographic data, providing a framework for the assessment of differences in both mean concentration and laminar distribution patterns across multiple subjects. Initially, laminar receptor distribution patterns for cortical areas are quantified by sampling density profiles in a series of regions of interest (ROIs) from digitalized autoradiographs and computing a mean profile per ROI. These ROI mean profiles are then corrected for distortions in the laminar pattern introduced by cortical folding and averaged to yield a mean profile per area, receptor and hemisphere. Differences in mean binding site concentration between areas are quantified by the asymmetry coefficient which is the difference of the mean concentrations divided by their sum. To quantify differences in laminar receptor distribution patterns between areas, the effects of absolute binding site concentration are first removed by dividing each profile by its mean value. Differences in the laminar pattern were then quantified by calculating the Euclidean distance between these mean corrected profiles. For single subject analysis, we propose a permutation test, comparing the differences between the mean profiles for two areas to differences between groups of profiles randomly assembled from all ROIs sampled in either area. Group inference can then be based on a between-subject conjunction analysis after correcting p-values to control for multiple comparisons. The feasibility of the presented approach is demonstrated by an exemplary analysis of the neurochemical differences between the ventral parts of the second and the third visual area.