Contributions to auditory system conduction velocity: insights with multi-modal neuroimaging and machine learning in children with ASD and XYY syndrome

Abstract

Introduction: The M50 electrophysiological auditory evoked response time can be measured at the superior temporal gyrus with magnetoencephalography (MEG) and its latency is related to the conduction velocity of auditory input passing from ear to auditory cortex. In children with autism spectrum disorder (ASD) and certain genetic disorders such as XYY syndrome, the auditory M50 latency has been observed to be elongated (slowed).

Methods: The goal of this study is to use neuroimaging (diffusion MR and GABA MRS) measures to predict auditory conduction velocity in typically developing (TD) children and children with autism ASD and XYY syndrome.

Results: Non-linear TD support vector regression modeling methods accounted for considerably more M50 latency variance than linear models, likely due to the non-linear dependence on neuroimaging factors such as GABA MRS. While SVR models accounted for ~80% of the M50 latency variance in TD and the genetically homogenous XYY syndrome, a similar approach only accounted for ~20% of the M50 latency variance in ASD, implicating the insufficiency of diffusion MR, GABA MRS, and age factors alone. Biologically based stratification of ASD was performed by assessing the conformance of the ASD population to the TD SVR model and identifying a sub-population of children with unexpectedly long M50 latency.

Discussion: Multimodal integration of neuroimaging data can help build a mechanistic understanding of brain connectivity. The unexplained M50 latency variance in ASD motivates future hypothesis generation and testing of other contributing biological factors.

Keywords: MRI; Magnetoencephalography; XYY syndrome; autism spectrum disorder; brain; machine learning; pediatric.

Figure 1

(A) Sensor waveform of auditory evoked response. The auditory stimulus is presented at the vertical dotted line. The auditory M50 component is represented at the vertical solid line. (B) Topographic contour map of the MEG M50 response. (C) MEG M50 source dipoles in the left and right hemisphere and their (D) waveforms (left response on the top and right response on bottom). (E) Left and right regions of interest for DTI measurement of Heschl's gyrus. (F) Typical GABA MRS.

Figure 2

Relationship between FA and M50 in TD Linear Model 1.

Figure 3

Both FA and Age are significant predictors of M50 in TD Linear Model 2.

Figure 4

A subset of ASD subjects exhibited greater than 85% percentile later M50 latency than is predicted by the TD model of M50. This set of “outlier” ASD subjects comprises a biologically based subset which exhibited significantly lower GABA levels than the more model-conforming ASD subjects.

Figure 5

Comparison of 99% confidence interval (blue shaded region) for the Linear Model (R² = 0.52) and the SVR Model (R² = 0.82).

Figure 6

Factor importance compared across SVR models. Differences in Age and GABA sensitivity across groups indicate altered biological mechanisms modulate M50 latency in ASD and XYY syndrome. Noted differences in sensitivity between groups include Age and GABA. The ASD model has least sensitivity to AGE. The XYY syndrome model has enhanced sensitivity to GABA relative to the TD and ASD groups, suggesting a more important role for GABA in XYY syndrome.

Figure 7

A subset of ASD subjects exhibit greater than 99% percentile longer M50 than is predicted by the SVR TD model of M50 latency. This set of “outlier” ASD subjects comprises a biologically based subset with unknown biological basis.