Self-Supervised Speech Representations for Cross-Speaker Dysarthria Detection During Awake Craniotomy
Detecting intra-operative speech impairment during awake craniotomy is essential for preserving language function. However, automated detection remains challenging because operating-room recordings contain substantial acoustic interference, clinically relevant speech events are rare, and available cohorts are small and...
K. Chinmayi, Abdallah Nassib, Misy Harrison et al.
· 0 citations