Automated smartphone audiometry: Validation of a word recognition test app

Automated Forced-Choice Tests of Speech Recognition

PURPOSE

 This project was undertaken to develop automated tests of speech recognition, including speech-recognition threshold (SRT) and word-recognition test, using forced-choice responses and computerized scoring of responses. Specific aims were (1) to develop an automated method for measuring SRT for spondaic words that produces scores that are in close agreement with average pure-tone thresholds and (2) to develop an automated test of word recognition that distinguishes listeners with normal hearing from those with sensorineural hearing loss and which informs the hearing aid evaluation process.

METHOD

 An automated SRT protocol was designed to converge on the lowest level at which the listener responds correctly to two out of two spondees presented monaurally. A word-recognition test was conducted with monosyllabic words (female speaker) presented monaurally at a fixed level. For each word, there were three rhyming foils, displayed on a touchscreen with the test word. The listeners touched the word they thought they heard. Participants were young listeners with normal hearing and listeners with sensorineural hearing loss. Words were also presented with nonrhyming foils and in an open-set paradigm. The open-set responses were scored by a graduate student research assistant.

RESULTS

 The SRT results agreed closely with the pure-tone average (PTA) obtained by automated audiometry. The agreement was similar to results obtained with the conventional SRT scoring method. Word-recognition scores were highest for the closed-set, nonrhyming lists and lowest for open-set responses. For the hearing loss participants, the scores varied widely. There was a moderate correlation between word-recognition scores and pure-tone thresholds which increased as more high frequencies were brought into the PTA. Based on the findings of this study, a clinical protocol was designed that determines if a listener's performance was in the normal range and if the listener benefited from increasing the level of the stimuli.

CONCLUSION

 SRTs obtained using the automated procedure are comparable to the results obtained by the conventional clinical method that is in common use. The automated closed-set word-recognition test results show clear differentiation between scores for the normal and hearing loss groups. These procedures provide clinical test results that are not dependent on the availability of an audiologist to perform the tests.

[Application areas of medical apps in otolaryngology]

The use of medical apps is becoming increasingly important as it offers new solutions in healthcare. Steadily growing computing and storage capacities in combination with high-precision sensors make smartphones effective tools for medical diagnostics and treatment. The use of this technology offers immense advantages, such as direct availability or independence from opening times. However, it also harbors risks such as unfiltered data storage and transmission. The consulting physician should exercise great care when selecting and recommending apps, particularly since only a few have been certified as medical devices to date. There is a steadily growing range of products on the market for otorhinolaryngology. The scientific evidence and quality of the apps vary widely, but tools exist for their validation by physicians and patients. The present training course is intended to help increase knowledge in this new, rapidly developing area.

APP Medical Diagnostic Check-up Consultation System Based on Speech Recognition

Background:

Medical test orders can display the physiological functions of patients by using medical means. The medical staff determines the patient's condition through medical test orders and completes the treatment. However, for most patients and their families, there are so many terminologies in the medical test list and they are inconvenient to understand and query, which would affect the patients’ cognition and treatment effect. Therefore, it is especially necessary to develop a consulting system that can provide related analysis after getting medical test data.

Objective :

This paper starts with information acquisition and speech recognition. It proposes a natural scene information acquisition and analysis model based on deep learning, focusing on improving the recognition rate of routine test list and achieving targeted smart search to allow users to get more accurate personalized health advice.

Methods :

Based on medical characteristics, considering the needs of patients, this paper constructs an APP-based conventional medical test consultation system, using artificial intelligence and voice recognition technology to collect user input; analyzing user needs with the help of conventional medical information knowledge database.

Results:

This model combines speech recognition and data mining methods to obtain routine test list data and is suitable for accurate analysis of problems in routine check-up procedure. The app provides effective explanations and guidance for the treatment and rehabilitation of patients.

Conclusion:

It organically links the Internet with personalized medicine, which can effectively improve the popularity of medical knowledge and provide a reference for the application of medical services on the Internet. Meanwhile, this app can contribute to the improvement of medical standards and provide new models for modern medical management.

Development and evaluation of a tablet-based diagnostic audiometer

Objective: To develop and evaluate a software application capable of conducting Pure-Tone Audiometry tests in clinical practice. Design: We designed and developed a mobile software application for iPad devices that performs Pure-Tone Audiometry according to ANSI and IEC standards. The application is proposed to be operated by a trained audiologist inside a sound booth. No extra equipment is required. Hence, it updates the procedure by showing the versatility of the proposed system. Particularly, it provides manual and automated measurement, including air- and bone-conduction audiometry. Study sample: Twenty-nine participants-patients of Papageorgiou Hospital, Thessaloniki, Greece were tested, with all degrees of hearing sensitivity. Manual air- and bone-conduction Pure-Tone Audiometry was conducted inside a sound booth. Participants were tested with conventional audiometry and the audiometric application, in order to validate the tablet-based audiometer for measuring hearing thresholds. Results: The majority (90.9%) of air-conduction estimated hearing thresholds and (90.8%) of air-bone gaps were within 5 dB, compared to results obtained by conventional audiometry. Thus, threshold differences were not significant. Conclusions: The proposed audiometer is a reliable and valid tool for hearing assessment. Owing to certain limitations, mobile devices can provide a feasible substitute for conventional audiometry in clinical practice.

Automated Smartphone Audiometry: A Preliminary Validation of a Bone-Conduction Threshold Test App

Objective:

To develop and validate an automated smartphone app that determines bone-conduction pure-tone thresholds.

Methods:

A novel app, called EarBone, was developed as an automated test to determine best-cochlea pure-tone bone-conduction thresholds using a smartphone driving a professional-grade bone oscillator. Adult, English-speaking patients who were undergoing audiometric assessment by audiologists at an academic health system as part of their prescribed care were invited to use the EarBone app. Best-ear bone-conduction thresholds determined by the app and the gold standard audiologist were compared.

Results:

Forty subjects with varied hearing thresholds were tested. Sixty-one percent of app-determined thresholds were within 5 dB of audiologist-determined thresholds, and 79% were within 10 dB. Nearly all subjects required assistance with placing the bone oscillator on their mastoid.

Conclusion:

Best-cochlea bone-conduction thresholds determined by the EarBone automated smartphone audiometry app approximate those determined by an audiologist. This serves as a proof of concept for automated smartphone-based bone-conduction threshold testing. Further improvements, such as the addition of contralateral ear masking, are needed to make the app clinically useful.