Research Article: Predicting cochlear dead regions in patients with hearing loss through a machine learning-based approach: A preliminary study

Date Published: June 3, 2019

Publisher: Public Library of Science

Author(s): Young-Soo Chang, Heesung Park, Sung Hwa Hong, Won-Ho Chung, Yang-Sun Cho, Il Joon Moon, Andreas Buechner.


We propose a machine learning (ML)-based model for predicting cochlear dead regions (DRs) in patients with hearing loss of various etiologies. Five hundred and fifty-five ears from 380 patients (3,770 test samples) diagnosed with sensorineural hearing loss (SNHL) were analyzed. A threshold-equalizing noise (TEN) test was applied to detect the presence of DRs. Data were collected on sex, age, side of the affected ear, hearing loss etiology, word recognition scores (WRS), and pure-tone thresholds at each frequency. According to the cause of hearing loss as diagnosed by the physician, we categorized the patients into six groups: 1) SNHL with unknown etiology; 2) sudden sensorineural hearing loss (SSNHL); 3) vestibular schwannoma (VS); 4) Meniere’s disease (MD); 5) noise-induced hearing loss (NIHL); or 6) presbycusis or age-related hearing loss (ARHL). To develop a predictive model, we performed recursive partitioning and regression for classification, logistic regression, and random forest. The overall prevalence of one or more DRs in test ears was 20.36% (113 ears). Among the 3,770 test samples, the overall frequency-specific prevalence of DR was 6.7%. WRS, pure-tone thresholds at each frequency, disease type (VS or MD), and frequency information were useful for predicting DRs. Sex and age were not associated with detecting DRs. Based on these results, we suggest possible predictive factors for determining the presence of DRs. To improve the predictive power of the model, a more flexible model or more clinical features, such as the duration of hearing loss or risk factors for developing DRs, may be needed.

Partial Text

A cochlear dead region (DR) is defined as a region in the cochlea where the inner hair cells (IHCs) and/or neurons lose normal function at a related frequency. Detecting the presence of DRs is important in clinical practice. A previous study had reported that DRs are associated with potentially poor hearing thresholds on follow-up audiograms in patients with sudden sensorineural hearing loss (SSNHL) [1]. Since it is debatable whether the presence of DRs, especially in high frequencies, is associated with hearing aid fitting and amplification selection [2–6], studies to detect the presence of DRs and to reveal their role continue.

The overall prevalence of one or more DRs in the test ears evaluated using the TEN (HL) test was 20.36% (113 ears). Among the 3,770 test samples, the overall frequency-specific prevalence of DR was 6.7%. The mean age (± standard deviation) of the study population was 56.4 ± 13.8 years. Two hundred and five patients were tested in one ear, and 175 patients were tested in both ears. Descriptive statistics of the study population are listed in Table 1.

Detection of DRs is the first step in understanding the clinical importance of DR. One of the biggest hurdles for evaluating the role of DRs is that the TEN (HL) test is very time consuming. Although previous studies have revealed reliable indicators of DRs based on detection by TEN (HL) tests [2, 11, 12], the prevalence and possible indicators of DRs differ according to the study population [2, 12, 25, 26]. Therefore, it is still unclear which patients beyond those with severe-to-profound hearing loss should undergo the TEN (HL) tests. There have been no previous studies that specifically address DR prediction as a function of frequency-specific information.




Leave a Reply

Your email address will not be published.