Case Report: Tailored ASR for Dysarthria and Global Aphasia

A single-case proof-of-concept study evaluates a personalized automatic speech recognition system for a patient with severe dysarthria and global aphasia. Davide Mulfari authored the report, which was last updated on June 3, 2026. The study compares the accuracy of a speaker-dependent VIVOCA system against rehabilitation professionals.

Use Cases

• Benchmarking ASR performance against human listeners based on the described accuracy comparison
• Training speaker-dependent models for atypical speech based on the described method of recording 1,120 utterances
• Evaluating prompting modalities for speech data collection based on the five modalities mentioned

Strengths

• Report includes specific performance metrics: ASR accuracy of 72.65% and human listener mean accuracy of 56.75%
• Study design details are provided: 1,120 training utterances and 936 independent evaluation utterances

Limitations

• Row count is unknown, which may limit suitability assessment
• Column-level documentation is absent; field semantics must be inferred after download

Provenance

Source

Davide Mulfari

Collection Method

A single-case proof-of-concept study involving recorded speech samples and a trained convolutional neural network.

Freshness

Last updated 2026-06-03 04:47:19; freshness should be verified