White matter correlates of face recognition impairments in cerebral visual impairment
Poster No:
2197
Submission Type:
Abstract Submission
Authors:
Marie Drottar1, Negin Nadvar2, Claire Manley3, Peter Bex4, Daniel Dilks5, John Ravenscroft6, Corinna Bauer2
Institutions:
1Massachusetts General Hospital, Boston MA, 2Massachusetts General Hospital, Boston, MA, 3Schepens Eye Research Institute,Mass Eye and Ear Infirmary, Boston, MA, 4Northeastern University, Boston, MA, 5Emory University, Atlanta, GA, 6University of Edinburgh, Edinburgh, United Kingdom
First Author:
Co-Author(s):
Introduction:
Children with early brain injury are at increased risk for cerebral visual impairment (CVI). CVI is a leading cause of pediatric visual impairment, with heterogeneous manifestations that may include impaired visual function (i.e., reduced visual acuity) and higher order visual perceptual dysfunctions. Previous research from our lab indicates that face recognition may also be impaired in those with CVI (Bauer et al., Vision 2023). However, the neural correlates remain unclear. In this work, we investigated the white matter structural changes of key white matter tracts associated with the ventral visual stream, specifically the optic radiations (OR), inferior longitudinal fasciculus (ILF), and the interior fronto-occipital -fasciculus (IFOF). As a secondary analysis, we investigated the relationship between tract volume and face recognition impairments in individuals with CVI.
Methods:
Multishell diffusion data (97 direction, b = 500, 1000, 2000, 3000 s/mm2 was acquired for seven participants with CVI (22.5 years, 3.53 s.d., 3 females) and seven controls (20.67 years, 4.92 s.d,, 4 females) on a 3T Phillips Ingenia Elition X. Preprocessing of diffusion data included top-up, brain extraction, motion, and eddy correction using FSL functions. Whole brain tractograms were reconstructed using MRTRIX3 default parameters (Tournier et al., 2019). Tract segmentation of individual tracts was done with TractSeg (Wasserthal et al., 2018), which segments white matter tracts using constrained spherical deconvolution (CSD) to extract three dominant diffusion directions in each voxel. Tract volumes were calculated and visualized in TrackVis.
The face recognition task was completed by each of the seven CVI and a subset of the control participants (n=2) (Bauer et al.,Vision, 2023). We used partial Spearman correlations to investigate the relationship between volume of the three tracts and performance on the face recognition task (threshold and proportion correct), while adjusting the potential effects of age. Bonferroni correction for multiple comparisons was used (critical p = 0.0042).
The face recognition task was completed by each of the seven CVI and a subset of the control participants (n=2) (Bauer et al.,Vision, 2023). We used partial Spearman correlations to investigate the relationship between volume of the three tracts and performance on the face recognition task (threshold and proportion correct), while adjusting the potential effects of age. Bonferroni correction for multiple comparisons was used (critical p = 0.0042).
Results:
Volume of the left ILF was significantly reduced in CVI compared to controls after adjusting for age and intracranial volume (F(2,13) = 14.13, p = 0.0032).
Significant correlations were observed between threshold and volume of the right IFOF (r = -0.95, p = 0.0042), whereby smaller volumes were associated with worse performance (i.e., increased threshold). There was a trend for a negative relationship with volume of the left ILF (r = -0.96, p = 0.0059) and right OR (r = -0.90, p = 0.014), but these did not survive correction for multiple comparisons. Similarly, proportion correct was significantly correlated with volume of the right IFOF (r = 0.99, p = 0.0001), whereby smaller volumes were associated with fewer correct responses. Trends were observed for the left ILF (r = 0.93, p = 0.0072) and right OR (r = 0.84, p = 0.038), however these did not survive multiple comparisons correction.
Significant correlations were observed between threshold and volume of the right IFOF (r = -0.95, p = 0.0042), whereby smaller volumes were associated with worse performance (i.e., increased threshold). There was a trend for a negative relationship with volume of the left ILF (r = -0.96, p = 0.0059) and right OR (r = -0.90, p = 0.014), but these did not survive correction for multiple comparisons. Similarly, proportion correct was significantly correlated with volume of the right IFOF (r = 0.99, p = 0.0001), whereby smaller volumes were associated with fewer correct responses. Trends were observed for the left ILF (r = 0.93, p = 0.0072) and right OR (r = 0.84, p = 0.038), however these did not survive multiple comparisons correction.
·Figure 1 A: All participants’ tractograms for left and right hemisphere ILF, IFOF, and OR B: Right hemisphere sagittal views C: Control
Conclusions:
These results indicate that there are specific changes in volume of key tracts involved with face processing in individuals with CVI. Further, these reduced volumes may be associated with impairments in face recognition frequently observed in this population. Additional analyses are warranted in a larger sample to verify the findings.
Disorders of the Nervous System:
Neurodevelopmental/ Early Life (eg. ADHD, autism)
Higher Cognitive Functions:
Higher Cognitive Functions Other
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 2
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
White Matter Anatomy, Fiber Pathways and Connectivity 1
Perception, Attention and Motor Behavior:
Perception: Visual
Keywords:
Data analysis
MRI
Pediatric Disorders
Plasticity
STRUCTURAL MRI
Tractography
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
Provide references using author date format
Tournier, J.-D., et al (2029), MRtrix3: A fast,flexible and open software framework for medical image processing and visualisation., Neuroimage, 202.
Wasserthal, J., et al, (2018) TractSeg--Fast and accurate white matter tract segmentation.n Neuroimage, 183.