Related Experiment Video
Updated: Jun 5, 2025
07:13
Modified Experimental Conditions for Noise-Induced Hearing Loss in Mice and Assessment of Hearing Function and Outer Hair Cell Damage
Published on: February 10, 2023
2.2K
Optimizing a Classification Model to Evaluate Individual Susceptibility in Noise-Induced Hearing Loss:
Shiyuan Li1,2,3, Xiao Yu1,2,3, Xinrong Ma1,2,3
1Department of Otolaryngology-Head and Neck Surgery, Shanghai Sixth People's Hospital, Shanghai Jiao Tong University School of Medicine, 600 Yishan Road, Shanghai, 200030, China, 86 18060587551.
JMIR Public Health and Surveillance
|December 4, 2024
Summary
A machine learning model accurately identifies noise-induced hearing loss (NIHL) susceptibility using specific hearing frequencies. This helps prevent hearing damage in noise-exposed workers by identifying those most at risk.
Area of Science:
- Occupational Health
- Audiology
- Biomedical Engineering
Background:
- Noise-induced hearing loss (NIHL) is a significant health issue, particularly among young adults in noisy occupations.
- Individual susceptibility to NIHL varies greatly, necessitating objective identification methods.
- Preventing severe NIHL requires identifying individuals highly sensitive to occupational noise exposure.
Purpose of the Study:
- To develop and validate an optimal model for classifying individuals as susceptible or resistant to NIHL.
- To explore specific phenotypic traits associated with NIHL susceptibility and resistance profiles.
Main Methods:
- Collected cross-sectional data from shipyard workers in Shanghai (2015-2021).
- Evaluated six classification methods for NIHL susceptibility and resistance.
- Developed a machine learning (ML)-based diagnostic model using hearing frequencies (0.25-12 kHz) to identify reliable predictors.
- Optimized the model using hearing thresholds (HTs) at 4 and 12.5 kHz.
Main Results:
- The ML model demonstrated high performance in classifying NIHL-susceptible (NIHL-SG) and NIHL-resistant (NIHL-RG) groups (accuracy=0.78, AUC=0.81).
- Younger age, shorter noise exposure duration, and lower cumulative noise exposure characterized the NIHL-SG.
- NIHL-SG individuals exhibited significantly higher HTs at 4 and 12.5 kHz compared to the NIHL-RG.
Conclusions:
- An ML-based diagnostic model utilizing mean HTs at 4 and 12.5 kHz is a reliable method for identifying NIHL susceptibility.
- This model can aid in targeted prevention strategies for workers at high risk of NIHL.
- Further research into genetic factors influencing NIHL susceptibility is warranted.
