Accuracy of AI Laryngeal Disorder Detection - Takeaways - MDSpire
Advertisement
Accuracy of AI Laryngeal Disorder Detection
Review of 88 studies found AI systems achieved high accuracy for identifying abnormal voices, but performance declined among higher-level laryngeal disorder classifications.
AI systems show high accuracy in detecting abnormal voices but lower performance in identifying specific laryngeal disorders.
2
Reported accuracies for distinguishing healthy from pathologic voices ranged from 88% to 99%, while specific disorder identification remained below 75%.
3
The review analyzed 88 studies evaluating AI approaches for laryngeal disorder detection, highlighting a decline in performance from detection to diagnosis.
4
Methodologic concerns included reliance on limited databases and sustained-vowel recordings, affecting the generalizability of AI performance.
5
Current evidence supports AI as a tool for screening and monitoring, but it cannot replace endoscopic assessment for specific diagnoses.
