Developing an Improvised Generative Speech Protocol for People with Aphasia: Music Enriched Verb Network Strengthening Treatment (MeVNeST)

Abstract

Music-based interventions (MBI) may facilitate communication outcomes for people with aphasia (PWA; Cheever et al., 2018; Magee et al., 2017). There are few theory-based, protocolized MBIs for PWA. This study developed a transdisciplinary, theory-based, highly structured improvised singing protocol, the Music enhanced Verb Network Strengthening Treatment (MeVNeST), that is designed to enhance the Verb Network Strengthening Treatment (VNeST) (Edmond, 2016). The feasibility and the preliminary outcomes of MeVNeST were evaluated. The study involved three stages: 1) protocol development; 2) protocol testing; 3) microanalysis. In Stage 1 Standard VNeST procedures were adapted to incorporate music components based on the theoretical premises for music and language in the brain, theories of musical expectation (Lerdahl & Krumhansl, 2007), and two frameworks for music-based intervention protocol development (Robb et al., 2011; Hanson-Abromeit, 2015). The outcome was a preliminary MeVNeST protocol. In Stage 2, two phases were implemented to evaluate the effects of MeVNeST compared with VNeST, and the effect of MeVNeST alone on word retrieval and sentence production. Effect sizes were calculated for probe scores. One participant who received seven weeks of MeVNeST showed improvements in the sentence production score for the VNeST probes of the trained verbs (d=3.22). No changes were observed on the Boston Naming Test, the Verb Naming Test, and the Western Aphasia Battery and the adjective control task. Other participants did not achieve significant changes in word retrieval and sentence production. This indicates the increased duration of MeVNeST in Phase 2 was necessary to produce the desired outcomes. In stage 3, the microanalysis results revealed that patterns of music interaction, called Patterns of Musical Influence (PMI), underpinned how the key musical components in the therapists’ improvised music influenced participants’ improvised singing. Musical strategies named Structuring Strategies and Cueing Strategies clarified the therapists’ musical strategies to facilitate participants’ performance of language tasks that were difficult. The MeVNeST protocol was revised based on the findings. This study suggests that MeVNeST is a feasible treatment approach that warrants further research. Future research is needed to test the MeVNeST protocol with a larger sample, and to optimize the capacity of improvised music to support the structured language task and escalate the generative process in the training of the semantic network.