Multimodal Neuroimaging Integration using fMRI and MEG in Youth with Autism Spectrum Disorder
- Wilkinson, Molly
- Advisor(s): Müller, Ralph-Axel
Abstract
Autism spectrum disorder (ASD) is a neurodevelopmental disorder that has been extensively studied using functional magnetic resonance imaging (fMRI). Studies have been mostly unimodal, with inconsistent findings. Combining fMRI with magnetoencephalography (MEG) is critical in accounting for dynamic aspects of brain activity, taking advantage of the high spatial vs. temporal resolution of each modality. The three-paper dissertation compares fMRI and MEG data from 12- to 21-year-olds with and without ASD. Study 1 (Wilkinson et al., 2022) compared fMRI BOLD signal and MEG event-related theta power in 33 ASD and 23 typically developing (TD) youth during a lexicosemantic task. Greater theta power was seen in the ASD group across several regions, whereas the BOLD signal was greater for the ASD group only in the anterior cingulate. No significant correlations were found between BOLD signal and theta power. Study 2 examined the relationship of N250m and N400m with fMRI BOLD signal in 30 ASD and 20 TD youth during lexicosemantic performance. The high-performing ASD subgroup had greater N250m in the inferior frontal region, and more bilateral N400m in inferior frontal and temporal regions. The low-performing ASD subgroup showed decreased N400m. fMRI BOLD signal did not show any significant group differences. Positive correlations between fMRI and MEG in multiple regions were seen predominantly in the low-performing ASD subgroup. Study 3 examined the relationship between two measures of the excitation:inhibition (E:I) ratio in adolescents with and without ASD. The Hurst exponent (HE) was estimated from resting-state fMRI data (34 ASD, 28 TD) and was found to be mostly increased in the ASD group. The aperiodic exponent (AE) was measured from resting-state MEG data (33 ASD, 22 TD) and was found to be greater in the ASD group across many brain regions. However, there was no clear relationship between fMRI HE and MEG AE. Overall findings from this dissertation indicate no clear relationship between fMRI and MEG, suggesting the two modalities are measuring distinct, yet complementary neural processes.