ARRS 2022 Abstracts

RETURN TO ABSTRACT LISTING


E1124. Neuroimaging of Autism Spectrum Disorder: A Review
Authors
  1. Faranak Rafiee; Johns Hopkins Hospital
  2. Roya Rezvani Habibabadi; Johns Hopkins Hospital
  3. Mina Motaghi; Johns Hopkins Hospital
  4. David Yousem; Johns Hopkins Hospital
  5. Ilyssa Yousem; Johns Hopkins Hospital
Background
Autistic Spectrum Disorder (ASD) is a neuropsychiatric continuum of disorders characterized by persistent deficits in social communication and restricted repetitive patterns of behavior that impede optimal functioning. Early detection and intervention in ASD children can mitigate the deficits in social interaction and result in a better outcome. Various non-invasive imaging methods and molecular techniques have been developed for the early identification of ASD characteristics. There is no general consensus on specific neuroimaging features of autism; however, quantitative MRI techniques have provided valuable structural and functional information in understanding the neuropathophysiology of ASD and how the autistic brain changes during childhood, adolescence, and adulthood.

Educational Goals / Teaching Points
Goals include: clinical manifestations of ASD (DMS-5 classification and Neuropsychiatric tests used to monitor/diagnose the disease); presumed pathophysiology of ASD; role of structural imaging in monitoring/diagnosis of ASD (conventional MRI, Volumetric methods, diffusion tensor imaging [DTI]); role of functional imaging in monitoring/diagnosis of ASD (task based functional MRI [fMRI], resting state fMRI [rs-fMRI], PET scanning, MR spectroscopy [MRS]); and role of artificial intelligence (AI) in monitoring/diagnosis of ASD.

Key Anatomic/Physiologic Issues and Imaging Findings/Techniques
Studies highlight the exaggerated synaptic pruning, anomalous gyrification, interhemispheric under- and overconnectivity and excitatory glutamate and inhibitory GABA imbalance theories of ASD. However, the literature regarding neuroimaging of ASD is confusing because of the impurity of the study subjects, the gross changes in the brain’s structure and function during the age range when ASD becomes manifest, and the superimposition of environmental factors and other neuropsychological disorders, all of which may skew the results. Therefore the neurosciences community would benefit most from a longitudinal study of children from the onset of suspected ASD between ages 3 and 6 years and then serially at 3 - 5 year intervals using the imaging modalities that are most immune to compliance and cost issues. This means collecting data sets of 3D volumetry, DTI, rs-fMRI, and GABA and glutamine/glutamate MRS in compliant (or sedated) children and comparing them to similarly handled TD controls longitudinally.

Conclusion
The neuroimaging studies to date have consolidated around the theories of (MRS) excitation-inhibition dysregulation, (DTI) white matter disconnectome, (voxel-based morphometry, 3D MRI) frontotemporal brain dysmorphism, and (fMRI) over and under activation of various brain regions during task-based studies and underconnectivity on resting state studies. As machine learning incorporates more structural and functional components to the AI analysis, neuroimaging will play a greater role in identifying the autism cerebral functional phenotype. This will help clinicians and behavioral therapists in those frequent instances when there may be a variety of neuropsychological disorders that co-exist or are considered in children with learning disabilities.