Researchers at Asan Medical Center (AMC) have developed artificial intelligence (AI)-based prediction models of cochlear implant surgery result, the hospital said Tuesday.
|Professors Chung Jong-woo and Professor Park Hong-ju|
The number of adult hearing loss patients is increasing in Korea as the population ages. Such symptoms have raised the need for a platform that can provide rapid diagnoses, proper treatments, and the rehabilitation of hearing loss for patient’s to live a social life and have a healthy mental status.
As researches indicate that hearing loss is closely related to dementia recently, the need for a proper platform has increased even more.
Patients experiencing hearing loss usually go through cochlear implantation, which directly stimulates the nervous system, as an alternative for hearing-impaired patients who cannot understand speech even if they use a hearing aid. However, there are individual differences in the results of surgery, which, in turn, leads to patients pondering over whether or not to receive the operation.
To solve such problems, the research team, led by Professors Chung Jong-woo and Park Hong-ju of the department of otorhinolaryngology at the hospital and researchers from the University of Southern California, has developed an artificial intelligence algorithm that predicts surgical outcomes by analyzing preoperative data from adult patients with cochlear implants.
The new platform showed 95 percent accuracy compared with the real world data of 120 patients who previously underwent cochlear implantation at AMC.
The artificial intelligence algorithm used by the research team is a nonlinear machine-learning method using preoperative variables such as the duration of hearing loss, the presence of residual hearing, and the length of hearing aid use. The technique overcomes the limits of the existing prediction model, which provided only analysis of the preoperative factors only in a linear way. The accuracy of the current prediction model was around 67 percent.
Also, with the development of a predictive model of the cochlear implant surgery, medical workers can speed up patients’ recovery by turning on the cochlear implant at the appropriate time after the operation as adult patients who have already heard the sound can quickly adapt to the stimulation of the sound through the cochlear implants.
“With the development of this AI predictive model, we hope to raise the satisfaction of the patients by providing them with realistic expectations before the operation,” Professor Chung said. “The platform will also provide appropriate hearing rehabilitation training to the patients.”
<© Korea Biomedical Review, All rights reserved.>