A deep learning approach for acoustic-based identification of muscle tension dysphonia and spasmodic dysphonia - Takeaways - MDSpire

A deep learning approach for acoustic-based identification of muscle tension dysphonia and spasmodic dysphonia

  • By

  • Zhou Zhou

  • Yuan Cheng

  • Qingyi Ren

  • Yike Li

  • Xu Yuanyue

  • Jing Kang

  • Cheng Lu

  • Pingjiang Ge

  • July 9, 2026

  • 0 min

Share

  • 1

    The study developed a deep learning model to differentiate between healthy voices, spasmodic dysphonia (SD), and muscle tension dysphonia (MTD) using voice recordings.

  • 2

    The AI model achieved an accuracy of 89.5% in distinguishing healthy voices from disordered ones and 71.6% in ternary classification of healthy, MTD, and SD.

  • 3

    The model's performance surpassed that of human experts, who had average accuracies of 78.2% in binary and 60.6% in ternary classifications.

  • 4

    The study highlights the diagnostic challenges in differentiating SD and MTD due to their overlapping perceptual features and similar laryngoscopic findings.

  • 5

    This research introduces a novel three-class classification architecture that enhances the diagnostic capabilities for voice disorders in clinical settings.

Original Source(s)

A deep learning approach for acoustic-based identification of muscle tension dysphonia and spasmodic dysphonia

  • by Zhou Zhou, Yuan Cheng, Qingyi Ren, Yike Li, Xu Yuanyue, Jing Kang, Cheng Lu, Pingjiang Ge

  • July 9, 2026

Related Content

Patients' perception towards large language models in otorhinolaryngology, head and neck surgery: a single-centre survey

Comparison of the effectiveness of autofluorescence devices in the detection of parathyroid glands during thyroid operations: a propensity score matched analysis (PSM)

How Does Social Media Influence Health Decisions?

National survey findings suggest many US adults report making health decisions based on social media despite widespread concerns about the accuracy of health information shared to the platforms.