Iterative LLM-based improvement for French Clinical Interview Transcription and Speaker Diarization
For researchers working on automatic speech recognition in medical settings, this work offers a practical LLM-based post-processing method to improve transcription accuracy and speaker attribution, though it is incremental and limited to small datasets.
This study proposes a multi-pass LLM post-processing architecture for French clinical interview transcription and speaker diarization, achieving significant WDER reductions on suicide prevention conversations (p<0.05, n=18) with zero output failures and RTF 0.32.
Automatic speech recognition for French medical conversations remains challenging, with word error rates often exceeding 30% in spontaneous clinical speech. This study proposes a multi-pass LLM post-processing architecture alternating between Speaker Recognition and Word Recognition passes to improve transcription accuracy and speaker attribution. Ablation studies on two French clinical datasets (suicide prevention telephone counseling and preoperative awake neurosurgery consultations) investigate four design choices: model selection, prompting strategy, pass ordering, and iteration depth. Using Qwen3-Next-80B, Wilcoxon signed-rank tests confirm significant WDER reductions on suicide prevention conversations (p<0.05, n=18), while maintaining stability on awake neurosurgery consultations (n=10), with zero output failures and acceptable computational cost (RTF 0.32), suggesting feasibility for offline clinical deployment, pending validation on larger corpora.