Auditory-Motor Mapping Training and Language-Related Pathways in Minimally Verbal Children with ASD

Background: Although up to 30% of individuals with ASD are minimally verbal, extremely few interventions can reliably produce improvements in speech output. Recently, we developed Auditory-Motor Mapping Training (AMMT), a novel intonation-based intervention, which aims to facilitate speech output in minimally verbal children with ASD (Wan et al., 2010, 2011). This intervention involves the mapping of sounds to articulatory actions through intonation and bimanual rhythmic motor activities. AMMT is built upon the musical strengths and preferences that have been observed in these children. Furthermore, associating sounds with actions engages an auditory-motor network of brain regions important for speech that has been reported to be dysfunctional in ASD.

Objectives: The overall aim of the study is to examine the efficacy of AMMT in facilitating speech output in minimally verbal children with ASD. We first conducted a proof-of-concept study to determine if AMMT can result in significant speech improvements (Wan et al., 2011). After establishing proof-of-concept, we are now in the process of comparing the efficacy of AMMT with a Control Therapy (CT). In addition, we also collected diffusion tensor imaging (DTI) data in some of these minimally verbal children, to determine if their language-related pathways (arcuate fasciculus, uncinate fasciculus) are abnormal. 

Methods: Treatment study - 6 minimally verbal children with ASD who had no words participated in the proof-of-concept AMMT study. So far, 10 additional minimally verbal children have participated in the AMMT vs. CT study. These children undergo intensive one-on-one treatment sessions 5 times/week for a total of 25 sessions. In AMMT, a pair of tuned drums is used, and the speech therapist introduces the target words by intoning the words while simultaneously tapping the drums to facilitate sound-motor mapping. In CT, the key components of AMMT (intonation and hand-motor actions) are omitted, but CT is also designed to promote speech production. All children are assessed on their consonant-vowel productions multiple times before, during, and after therapy. Imaging study – Minimally verbal children with ASD and their matched controls (N=8) participated. Diffusion tensors were calculated for every voxel using FSL, and tractography was then applied to the DTI data to reconstruct white matter tracts. The two language-related tracts of interest were the arcuate fasiculus and the uncinate fasciculus.

Results: Results from the ongoing treatment study show that relative to CT, an intensive course of AMMT leads to significantly greater improvements in speech production, generalizing to items not trained during the therapy sessions. Results from the ongoing DTI study show that both language-related tracts are actually present in all minimally verbal children with ASD. Compared to those in typically developing children, however, the arcuate fasciculi of the minimally verbal children show reduced volumes and a right>left (rather than typical left >right) asymmetry. No such differences are observed in the uncinate fasciculus.

Conclusions: AMMT appears to have significant potential in facilitating speech output in minimally verbal children with ASD. Its effectiveness may lie in its ability to engage and potentially remodel the language-related pathway that is abnormal in these children.