iPhone Automated Audiogram Image Recognition Has Low Accuracy - Physician's Weekly

FRIDAY, Jan. 6, 2023 (HealthDay News) — The automated audiogram representation designation diagnostic successful the iPhone has debased accuracy, according to a probe missive published online Jan. 5 successful JAMA Otolaryngology-Head & Neck Surgery.

Janet S. Choi, M.D., from the University of Minnesota successful Minneapolis, and colleagues assessed the accuracy of the automated audiogram representation designation successful the iPhone utilizing 500 randomly selected audiogram reports from a wellness strategy objective proceeding database representing assorted degrees of proceeding loss.

The researchers recovered that the accuracy of the automated audiogram representation designation was 7.0 and 11.8 percent for close and near ears, respectively. For much than 80 percent of ears, the diagnostic was incapable to observe capable thresholds to cipher pure-tone average, prompting the idiosyncratic to manually participate the audiogram data. When proceeding was asymmetric, accuracy was importantly higher (odds ratios, 2.26 and 2.27 for the close receptor and near ear, respectively). Accuracy was not associated with type, grade of proceeding loss, oregon fig of missing thresholds.

“Future efforts volition beryllium needed to amended the accuracy of automated audiogram representation designation if this attack is utilized for the customization of aboriginal over-the-counter proceeding aids,” the authors write.

One writer disclosed fiscal ties to Apple and the regenerative medicine industry.

Abstract/Full Text (subscription oregon outgo whitethorn beryllium required)

Copyright © 2022 HealthDay. All rights reserved.