Statistical Analysis of fNIRS Data: Consideration of Spatial Varying Coefficient Model of Prefrontal Cortex Activity Changes During Speech Motor Learning in Apraxia of Speech

Abstract

Apraxia of speech is an impairment in the planning and programming of speech typically accompanied by aphasia (language impairment) secondary to a left hemisphere stroke. It is unknown if the structural and functional connections to the damaged area implicate the integrity of the cognitive functions of the prefrontal cortex (PFC). The present study examines the feasibility of measuring hemodynamic activity in the PFC in response to the structure of practice and during treatment. This multiple-baseline single case-design study involving two individuals with chronic acquired apraxia of speech measured the hemodynamic changes in PFC activity during treatment across the intervention period using functional near-infrared spectroscopy (fNIRS). Two models-a generalized linear model and a spatial varying coefficient model-are used to distinguish the repeated measures of PFC activity differences corresponding to the stage of practice and time of intervention. There were significant differences in the pattern of PFC activity associated with the structure of practice and the time of intervention. The outcomes from this pilot study demonstrate the utility of fNIRS to identify cognitive effort during speech motor learning. The implications include consideration for statistical methods used for fNIRS analysis and its potential use as a clinical tool to complement behavior changes to guide patient-directed intervention to optimize patient outcomes.

Case Report: A Case Study on the Neurodevelopmental Profile of a Child With Pallister–Killian Syndrome and His Unaffected Twin

“…Malformations of this area result in speech and language deficits that are often associated with CAS ( 29 ). Johnson et al ( 30 ) investigated the use of fNIRS to explore brain functioning in the prefrontal cortex during treatment for CAS. Though fNIRS does not provide anatomical information, our results map cortical activation up to 1-cm deep during language-based assignments.…”