Methodological Considerations in AI Predictions of Hearing Outcomes After Intact Canal Wall Mastoidectomy

Summary: This study evaluated the performance of a machine learning model and investigated the effects of major prognostic factors on hearing outcomes after ICW mastoidectomy with tympanoplasty. The results showed that preoperative hearing status was the most important factor, and artificial intelligence models were effective in predicting hearing outcomes.

Summary: This study aims to evaluate the correlation between the site, size, and duration of tympanic membrane perforation and hearing loss using pure tone audiogram and surgical outcomes. The results show a significant relationship between the size and site of the perforation with the degree of hearing loss, with larger perforations resulting in greater loss. Central perforations were associated with more significant hearing loss compared to posterior perforations, contradicting the assumption that the site and size of a tympanic membrane perforation do not affect the degree of conductive hearing loss. No correlation was found between the duration of the disease and the degree of hearing loss.