Self-Improved Language Production in Nonfluent Aphasia Through Automated Recursive Self-Feedback

Gerald C Imaezue,Ofer Tchernichovski,Mira Goral
DOI: https://doi.org/10.1044/2024_AJSLP-23-00320
2024-09-20
Abstract:Background: Persons with nonfluent aphasia (PWNA) use feedback from external agents (e.g., speech-language pathologists) and self-feedback to improve their language production. The extent to which PWNA can improve their language production using their self-feedback alone is underexplored. In a proof-of-concept study, we developed an automated recursive self-feedback procedure to demonstrate the extent to which two PWNA who used self-feedback alone improved their production of sentences from trained and untrained scripts. In the current study, we use the Rehabilitation Response Specification System as a framework to replicate our initial findings. Method: We tested the effects of two treatments: script production with recursive self-feedback and script production with external feedback in four persons with chronic nonfluent aphasia. We compared the effects of treatment by measuring percent script produced, speaking rate, and speech initiation latency of trained and untrained scripts. The participants received the treatments remotely through mini tablets using two versions of a mobile app we developed. All the participants received each treatment intensively for 14 sessions across 2-3 weeks. We estimated clinical improvements of production of sentences from trained and untrained scripts through nonoverlap of all pairs analysis of performance pretreatment and posttreatment. Results: Both treatments improved PWNA's language production. Recursive self-feedback improved speaking rate and speech initiation latency, which generalized to untrained scripts in all participants. External feedback treatment did not generalize to improvement in speaking rate in two participants. Conclusions: Our findings confirm our initial evidence that PWNA can self-improve their sentence production from scripts through recursive self-feedback. This novel procedure enables PWNA to autonomously enhance their language production over time. Given the evidence and the mechanics of the procedure, we propose that its utility is not constrained by linguistic idiosyncrasies across cultures. Consequently, it has the potential to bypass linguistic barriers to aphasia care. Supplemental material: https://doi.org/10.23641/asha.27007060.
What problem does this paper attempt to address?