Structural Features of the Mid Fusiform Sulcus in Autism Spectrum Disorders

Background:   A landmark paper published recently identified a minor sulcus bisecting the fusiform gyrus (FG) into lateral and medial subdivisions (Weiner et al., 2014). The “mid-fusiform sulcus” (MFS) was shown to be a stable characteristic in healthy individuals, present across ages and identifiable in right and left hemispheres of the brain. Furthermore, this macroanatomical feature was related to a microanatomical division between cytoarchitectonic regions FG1 and FG2 occurring at the fundus of the MFS in 90% of cases (Weiner et al., 2014).  Despite the importance of the FG in face processing and social cognition as well as the considerable literature demonstrating abnormal FG activation in ASD (Nomi & Uddin, 2015), no systematic investigation of MFS structure has been undertaken in an ASD population.

Objectives:   To characterize the anatomical organization of the MFS in children and adults with ASD and compare and contrast it with that in TD participants. 

Methods:   Structural MRI data from 49 subjects (23 ASD; 26 TD) collected on a Siemens 3T scanner were analyzed for this study. The sample consisted of 41 males and 8 females (4 ASD, 4 TD). Participant ages ranged from 9 to 40 years (ASD mean=21.13, TD mean=19.62) with no significant group differences [t(47)=.67, p=.51].  IQ ranged from 80 to 140 (ASD mean=113, TD mean=113) with no significant group differences [t(47)=.17, p=.68]. Each hemisphere was analyzed independently resulting in a total sample size of 98 hemispheres. Identification and classification of MFS in each hemisphere was performed using criteria established in Weiner et al. (2014). Criteria included 1) an “omega” shape on T1 coronal slices created by the deeper Collateral sulcus medially and Occipital-Temporal sulcus laterally; and 2) visible sulcus representation on the inflated brain surface. Identifiable MFS were classified into one of four surface patterns using Freesurfer 2D rendering and rough correspondence with an fsaverage brain MFS label created for this project. These patterns reflect a combination of fractionation and contiguity with surrounding sulci.      

Results:  The MFS was identifiable through inspection of the inflated surface and T1 image in 97% of hemispheres. ASD or TD group membership was not associated with unidentifiable MFS. Of the 95 hemispheres with identifiable MFS the following classification patterns emerged: 1) In both TD and ASD groups, unfractionated MFS patterns were slightly more common across hemispheres (56% ASD; 59% TD); 2) MFS which are independent of neighboring sulci are slightly more common across groups and hemispheres (53% ASD, 55% TD); 3) Classification of MFS gross anatomy was similar between ASD and TD groups (ASD: IA-14, IB-11, IIA-11, IIB-9; TD: IA-17, IB-10, IIA-13, IIB-10). 

Conclusions:   The MFS can be reliably identified in the ASD population. Its macroanatomical structure appears to follow largely similar patterns seen in TD individuals.  Planned follow-up analyses include addition of another 65 subjects to increase power as well as quantitative indices of sulcal length and depth measurement. Further elaboration of structural features of the MFS in ASD will be important for future investigations into structure-function relationships in this critical region of the social brain.