Disruptions of Fractional Amplitude of Low-Frequency Fluctuations in Autism Spectrum Disorder
Poster No:
429
Submission Type:
Abstract Submission
Authors:
Samuel Joseph1, Sheeba Anteraper2
Institutions:
1Austin Preparatory School, Reading, MA, 2UT Southwestern Medical Center, Dallas, TX
First Author:
Co-Author:
Introduction:
Whole-brain connectome-wide data-driven studies have reported disruptions in cerebrocerebellar intrinsic functional connectivity (FC) in young adults with high-functioning autism spectrum disorder (ASD) [1]. Detecting the fractional amplitude of low-frequency fluctuations (fALFF) [2] of the BOLD signal in the frequency window of interest can provide insights complementary to FC measures. To test this, we examined fALFF in a highly sampled (temporal resolution < 0.5s) resting-state fMRI dataset obtained from the Autism Brain Imaging Data Exchange (ABIDE II).
Methods:
Data Acquisition: We used the public dataset on ABIDE II (n=51; 16 ASD, 35 healthy controls) contributed by Michal Assaf, MD (Olin Neuropsychiatry Research Center), collected on 3T Siemens Skyra. Functional data (3mm voxels) had TR/TE/flip angle of 475ms/30ms/60°, multi-band factor 8, and 947 time-points. Anatomical data (0.8 mm voxels) had TR/TE/TI/flip angle of 2200ms/2.88ms/794ms/13°.
Data Analyses: CONN 22.a[3] and SPM12[4].
Preprocessing: A flexible pipeline [5] was used for realignment with correction of susceptibility distortion interactions, outlier detection, direct segmentation and MNI-space normalization, and smoothing (5 mm Gaussian kernel). Outlier scans were identified using ART [6] as acquisitions with framewise displacement above 0.5 mm or global BOLD signal changes above 3 standard deviations [7]. Functional and anatomical data were normalized into standard MNI space, segmented into grey matter, white matter, and CSF tissue classes, and resampled to 2 mm isotropic voxels following a direct normalization procedure using SPM unified segmentation and normalization [8] with the default IXI-549 tissue probability map template.
Denoising: This pipeline includes the regression of potential confounding effects characterized by white matter and CSF timeseries, motion parameters and their first order derivatives, outlier scans, session effects and their first order derivatives, and linear trends within each functional run, followed by bandpass filtering (0.008-0.09 Hz) of the BOLD timeseries. CompCor [9] noise components within white matter and CSF were estimated by computing the average BOLD signal as well as the largest principal components orthogonal to the BOLD average, motion parameters, and outlier scans within each subject's eroded segmentation masks.
First-level analysis: fALFF maps characterizing low-frequency BOLD signal variability at each voxel were estimated as the ratio between the root mean square of the BOLD signal after denoising and band-pass filtering, divided by the same measure computed before band-pass filtering [2].
Group-level analyses: For each individual voxel a separate General Linear Model was estimated, with first-level connectivity measures at this voxel as dependent variables (one independent sample per subject), and groups as independent variables. Voxel-level hypotheses were evaluated using multivariate parametric statistics with random-effects across subjects and sample covariance estimation across multiple measurements. Results for the between group analyses (ASD vs. healthy controls) were thresholded using a cluster-forming p < 0.005 (two-sided) voxel-level threshold, and a false discovery rate corrected p < 0.05 cluster-size threshold.
Data Analyses: CONN 22.a[3] and SPM12[4].
Preprocessing: A flexible pipeline [5] was used for realignment with correction of susceptibility distortion interactions, outlier detection, direct segmentation and MNI-space normalization, and smoothing (5 mm Gaussian kernel). Outlier scans were identified using ART [6] as acquisitions with framewise displacement above 0.5 mm or global BOLD signal changes above 3 standard deviations [7]. Functional and anatomical data were normalized into standard MNI space, segmented into grey matter, white matter, and CSF tissue classes, and resampled to 2 mm isotropic voxels following a direct normalization procedure using SPM unified segmentation and normalization [8] with the default IXI-549 tissue probability map template.
Denoising: This pipeline includes the regression of potential confounding effects characterized by white matter and CSF timeseries, motion parameters and their first order derivatives, outlier scans, session effects and their first order derivatives, and linear trends within each functional run, followed by bandpass filtering (0.008-0.09 Hz) of the BOLD timeseries. CompCor [9] noise components within white matter and CSF were estimated by computing the average BOLD signal as well as the largest principal components orthogonal to the BOLD average, motion parameters, and outlier scans within each subject's eroded segmentation masks.
First-level analysis: fALFF maps characterizing low-frequency BOLD signal variability at each voxel were estimated as the ratio between the root mean square of the BOLD signal after denoising and band-pass filtering, divided by the same measure computed before band-pass filtering [2].
Group-level analyses: For each individual voxel a separate General Linear Model was estimated, with first-level connectivity measures at this voxel as dependent variables (one independent sample per subject), and groups as independent variables. Voxel-level hypotheses were evaluated using multivariate parametric statistics with random-effects across subjects and sample covariance estimation across multiple measurements. Results for the between group analyses (ASD vs. healthy controls) were thresholded using a cluster-forming p < 0.005 (two-sided) voxel-level threshold, and a false discovery rate corrected p < 0.05 cluster-size threshold.
Results:
Results from fALFF analyses in ASD vs. healthy controls are shown in Fig. 1/Table 1. The two cerebellar clusters (bilateral Crus I and II), visualized on a flat map, and the cerebral cluster (left frontal pole) overlaid on a surface representation, overlap with the regions attributed to social cognition. There was no statistically significant difference in head-motion between the two groups.
Conclusions:
By leveraging a high-temporal resolution public dataset, and by using fALFF, a metric complementary to FC measures, we add to the growing body of evidence highlighting the role of cerebellum in autism [10]. Overall, our findings support the role of cerebrocebellar circuitry in brain function and dysfunction.
Disorders of the Nervous System:
Neurodevelopmental/ Early Life (eg. ADHD, autism) 1
Modeling and Analysis Methods:
Task-Independent and Resting-State Analysis 2
Keywords:
Autism
Cerebellum
Data analysis
FUNCTIONAL MRI
Open Data
1|2Indicates the priority used for review
Provide references using author date format
[2] Zou, Q. H., Zhu, C. Z., Yang, Y., Zuo, X. N., Long, X. Y., Cao, Q. J., ... & Zang, Y. F. (2008). An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: fractional ALFF. Journal of neuroscience methods, 172(1), 137-141.
[3] Nieto-Castanon, A. & Whitfield-Gabrieli, S. (2022). CONN functional connectivity toolbox: RRID SCR_009550, release 22. doi:10.56441/hilbertpress.2246.5840.
[4] Penny, W. D., Friston, K. J., Ashburner, J. T., Kiebel, S. J., & Nichols, T. E. (Eds.). (2011). Statistical parametric mapping: the analysis of functional brain images. Elsevier.
[5] Nieto-Castanon, A. (2020). FMRI minimal preprocessing pipeline. In Handbook of functional connectivity Magnetic Resonance Imaging methods in CONN (pp. 3–16). Hilbert Press.
[6] Whitfield-Gabrieli, S., Nieto-Castanon, A., & Ghosh, S. (2011). Artifact detection tools (ART). Cambridge, MA. Release Version, 7(19), 11.
[7] Power, J. D., Mitra, A., Laumann, T. O., Snyder, A. Z., Schlaggar, B. L., & Petersen, S. E. (2014). Methods to detect, characterize, and remove motion artifact in resting state fMRI. Neuroimage, 84, 320-341.
[8] Ashburner, J. (2007). A fast diffeomorphic image registration algorithm. Neuroimage, 38(1), 95-113.
[9] Behzadi, Y., Restom, K., Liau, J., & Liu, T. T. (2007). A component based noise correction method (CompCor) for BOLD and perfusion based fMRI. Neuroimage, 37(1), 90-101.
[10] Anteraper, S. A., Guell, X., Taylor, H. P., D'Mello, A., Whitfield-Gabrieli, S., & Joshi, G. (2019). Intrinsic Functional Connectivity of Dentate Nuclei in Autism Spectrum Disorder. Brain connectivity, 9(9), 692–702.