Machine Learning Models for the Hearing Impairment Prediction in Workers Exposed to Complex Industrial Noise: A Pilot Study.
Yanxia Zhao, Jingsong Li, Meibian Zhang and 4 others
PMID 30142102WHAT IT FOUND
Models tested in 1,113 workers could sort workers as hearing impaired or not, and one model could predict average noise-induced hearing loss in all but three factories.
This pilot is not ready for patient care.
Key findings
01The support vector machine model was reported as the best classifier, with an AUC of 0.808.
02The MLP regression model predicted mean noise-induced hearing loss in all but three factories.
03Age, exposure duration, equivalent A-weighted sound level, and median kurtosis were significantly associated with hearing loss.
STILL TO COME
How it was doneWhat they found
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
This is a pilot study aimed at demonstrating feasibility. The workers were from 17 factories in Zhejiang province, China, and only 1,113 of 1,644 subjects met inclusion criteria. Inclusion required no hearing protection use, so the results do not apply to workers who use protection. The regression model did not predict mean hearing loss in three factories, and the paper notes small factory samples or mobile work positions as possible reasons. The paper says more data and more risk factors are needed to improve performance.
Declared interests
The authors declare no conflicts of interest.
The easy way to misread this
Do not read these models as ready for hearing screening or patient care. This was a pilot study in Chinese factory workers, and the regression model did not predict mean hearing loss in three factories.