Detection of Pathological Voice Samples Using a Text-to-Speech–Inspired Approach With Multiple Model Architectures: Method Development and Comparative Evaluation Study - Takeaways - MDSpire
Coming Soon: Introducing MDSpire News. Learn more
Conexiant’s news site is now MDSpire News. Learn more

Identifying Pathological Voice Samples With a Text-to-Speech–Inspired Method: Development and Comparison of Multiple Model Architectures

  • By

  • Marie Sophie Walter

  • Muhittin Demir

  • Mattias Heinrich

  • Felix Nensa

  • Stephan Lang

  • Nadia Sadok

  • René Hosch

  • October 19, 2026

Share

  • 1

    Voice disorders have diverse functional, organic, and neurological causes, and overlapping acoustic features complicate automated classification.

  • 2

    Limited access to otolaryngology and phoniatrics services can delay diagnosis and therapy, motivating scalable screening approaches.

  • 3

    AI systems have performed well on curated voice datasets, but the 208-sample VOICED database covers only 4 diagnostic classes.

  • 4

    The Saarbrücken Voice Database includes 2225 speaker or recording-session entries across 72 diagnostic categories, reflecting greater clinical heterogeneity.

  • 5

    Prior SVD studies reported sensitivity and specificity of 0.82 and 0.83 with regression methods, while transformer ensembles achieved accuracy and area under the curve of 0.91.

Original Source(s)

Detection of Pathological Voice Samples Using a Text-to-Speech–Inspired Approach With Multiple Model Architectures: Method Development and Comparative Evaluation Study

  • by Marie Sophie Walter, Muhittin Demir, Mattias Heinrich, Felix Nensa, Stephan Lang, Nadia Sadok, René Hosch

  • October 19, 2026

Related Content

Dr. Jonathan Burke selected as director of the National Institute of Dental and Craniofacial Research

  • September 18, 2026

Robotic cochlear insertion shows early potential

In 50 procedures, surgeons achieved nearly universal full electrode placement without major complications, with robotic setup adding about 6 minutes.

Medical debt linked to higher late-stage cancer rates

Lung cancer showed the strongest absolute association, while prostate and breast cancer findings diverged from the broader pattern.