Audiometric Pattern in Moderate and Severe Tympanic Membrane Retraction

A Deep Learning Approach to Predict Conductive Hearing Loss in Patients With Otitis Media With Effusion Using Otoscopic Images

Importance

Otitis media with effusion (OME) is one of the most common causes of acquired conductive hearing loss (CHL). Persistent hearing loss is associated with poor childhood speech and language development and other adverse consequence. However, to obtain accurate and reliable hearing thresholds largely requires a high degree of cooperation from the patients.

Objective

To predict CHL from otoscopic images using deep learning (DL) techniques and a logistic regression model based on tympanic membrane features.

Design, Setting, and Participants

A retrospective diagnostic/prognostic study was conducted using 2790 otoscopic images obtained from multiple centers between January 2015 and November 2020. Participants were aged between 4 and 89 years. Of 1239 participants, there were 209 ears from children and adolescents (aged 4-18 years [16.87%]), 804 ears from adults (aged 18-60 years [64.89%]), and 226 ears from older people (aged >60 years, [18.24%]). Overall, 679 ears (54.8%) were from men. The 2790 otoscopic images were randomly assigned into a training set (2232 [80%]), and validation set (558 [20%]). The DL model was developed to predict an average air-bone gap greater than 10 dB. A logistic regression model was also developed based on otoscopic features.

Main Outcomes and Measures

The performance of the DL model in predicting CHL was measured using the area under the receiver operating curve (AUC), accuracy, and F1 score (a measure of the quality of a classifier, which is the harmonic mean of precision and recall; a higher F1 score means better performance). In addition, these evaluation parameters were compared to results obtained from the logistic regression model and predictions made by three otologists.

Results

The performance of the DL model in predicting CHL showed the AUC of 0.74, accuracy of 81%, and F1 score of 0.89. This was better than the results from the logistic regression model (ie, AUC of 0.60, accuracy of 76%, and F1 score of 0.82), and much improved on the performance of the 3 otologists; accuracy of 16%, 30%, 39%, and F1 scores of 0.09, 0.18, and 0.25, respectively. Furthermore, the DL model took 2.5 seconds to predict from 205 otoscopic images, whereas the 3 otologists spent 633 seconds, 645 seconds, and 692 seconds, respectively.

Conclusions and Relevance

The model in this diagnostic/prognostic study provided greater accuracy in prediction of CHL in ears with OME than those obtained from the logistic regression model and otologists. This indicates great potential for the use of artificial intelligence tools to facilitate CHL evaluation when CHL is unable to be measured.

Comparative analysis of moderate and severe tympanic membrane retractions in children and adults

OBJECTIVE

Owing to the limited literature demonstrating the correlation between the degree of severity of retractions and the degree of hearing loss in children and adults, the study aimed to compare the differences in the location, the severity, and the air-bone gap (ABG) of tympanic membrane (TM) retractions in children and adults.

METHODS

Cross-sectional study, in a tertiary hospital. Consecutive patients with moderate or severe TM retractions (661 ears) between August 2000 and January 2019 were evaluated. The average age (mean ± standard deviation) was 11.7 ± 3.3 years among pediatric patients (42.4%) and 46.4 ± 5 years among adults (57.6%). Video-otoscopy and pure tone audiometry were performed in all patients. The main outcome measures were the locations of retractions, their prevalence, and their severity; ABG thresholds measured at the 4-frequency pure-tone average (PTA).

RESULTS

The prevalence of pars flaccida (PF) retractions was higher in adults, while that of pars tensa (PT) was higher in children (p = 0.00). The degree of severity was similar between children and adults for isolated PF and PT retractions (p = 0.37 and p = 0.10, respectively). Effusion was similar in children (27.8%) and adults (33.3%). The median decibel hearing level (dB HL) (minimum-maximum) of the ABG PTA was 13.75 dB (0-57.5 dB HL) in children and 13.75 dB (0-58.7 dB) in adults (p = 0.48). There was no difference in the size of the ABG PTA between children and adults (p = 0.71), and in ABG size for isolated PF retractions (p = 0.14), PT retractions (p = 0.35), and association of PF and PT retractions (p = 0.56).

CONCLUSION

PT retractions were more prevalent in children and PF retractions in adults. There was no difference between the two groups based on the severity of the retraction. The size of the air-bone gaps was similar in children and adults.