Hearing-aid pre-selection through a neural network.

O Arntsen; H Koren; T Strøm

doi:10.3109/01050399609074964

Home/Archive/PubMed

PubMed · 8975998

Hearing-aid pre-selection through a neural network.

O Arntsen·H Koren·T Strøm

View original source ↗Export metadata ↓

Abstract

Selection of optimal hearing aids for a patient with a particular audiogram has become increasingly difficult as the number of available models on the market has grown considerably. In order to assist the hearing-aid fitter in this work, recent programming techniques using artificial neural networks have been investigated. Artificial neural networks are capable of recognizing audiograms and relating them to suitable hearing aids. A programme to assist in the early stage of hearing-aid selection was developed using these techniques.

Explore related subjects

Audiometry Correction of Hearing Impairment Diagnostic Errors Equipment Design Hearing Aids Hearing Disorders Humans Information Systems Nerve Net Norway Journal Article

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

O Arntsen, H Koren, T Strøm. 1996. Hearing-aid pre-selection through a neural network.. https://doi.org/10.3109/01050399609074964

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2006

An alternative strategy for universal infant hearing screening in tertiary hospitals with a high delivery rate, within a developing country, using transient evoked oto-acoustic emissions and brainstem evoked response audiometry.

OBJECTIVE: To formulate an alternative strategy for universal infants hearing screening in an Indian tertiary referral hospital with a high delivery rate, which could be extended to similar situations in other developing countries. The system should be able to diagnose, in a timely fashion, all infants with severe and profound hearing losses. METHODS: One thousand newborn were randomly selected. All underwent testing with transient evoked oto-acoustic emissions (TEOAE) in the first 48 hours of life. All TEOAE failures were followed up and repeat tests were performed at three weeks, three months and six months of age. Infants with acceptable TEOAE results at any of the four ages were discharged from the study. Infants with unacceptable TEOAE results at all the four ages underwent brainstem evoked response audiometry and oto-endoscopy. The 'pass rate' for TEOAE testing was calculated for all four ages. The time taken to perform TEOAE and brainstem evoked response audiometry was recorded for all subjects. These recordings were statistically analysed to find the most suitable strategy for universal hearing screening in our hospital. RESULTS: The pass rate for TEOAE was 79.0 per cent at < or =48 hours, 85.0 per cent at three weeks, 97.0 per cent at three months and 98.0 per cent at six months. The average time taken to perform the test was 12 minutes for TEOAE and 27 minutes for brainstem evoked response audiometry. Obstructed and collapsed external auditory canals were the two factors that significantly affected the specificity of TEOAE in infants < or =48 hours old. CONCLUSION: The concept of screening all neonates within the first 48 hours of life is impractical because the specificity of TEOAE is lowest at that age. Many false positive results are generated, such that a larger number must undergo brainstem evoked response audiometry, wasting time and resources. This can easily be avoided by delaying TEOAE screening until three months of age, when it has a substantially lower false positive outcome. We expect that implementation of this alternative strategy in our hospital will maximise the benefits of such a programme.

Audiometry↗

PubMed2006

The hearing profile among learners in schools in the Western Cape, South Africa.

The hearing profile amongst school learners in the Western Cape was investigated. The University of Cape Town's hearing-screening programme at 11 primary schools (n=1101) was audited and revealed a 13.8% referral rate amongst school-aged children. The possible effects of environmental noise on hearing-screening performance at 500Hz was investigated. When the hearing-screening results at 500Hz were excluded from the audit, the referral rate was reduced from 13.8% to 7.9%. The referral rate for hearing-screening failures amongst the learners in the sample was towards the high end of the range for developing countries (3.3-13.9%). More learners had unilateral than bilateral hearing failures while laterality, gender and geographical differences were minimal. This high referral rate suggests a strong need for hearing-screening of the school-aged population.

Audiometry↗

PubMed2006

Short-term results using Kurz titanium ossicular implants.

The efficiency of titanium middle ear prosthesis for ossicular reconstruction in chronic ear disease is investigated in a Scandinavian two-center retrospective study from a Norwegian tertiary otology referral center and a Finnish otology referral center. Retrospective chart reviews were performed for procedures involving 73 titanium prostheses between 1999 and 2004. All patients that underwent surgery including the Kurz Vario titanium prosthesis were included in the study, 38 procedures including the partial ossicular replacement prosthesis (PORP) and 35 procedures including the total ossicular replacement prosthesis (TORP). Mean follow-up was 14 months. The ossiculoplasty was performed alone (29 patients) or in combination with other chronic ear surgery procedures (34 patients). Comparisons of preoperative and postoperative pure tone averages (0.5, 1, 2, and 3 kHz) according to AAO-HNS guidelines are presented, as well as data for different PTA definitions. Otosurgery procedures, complications, revisions, and extrusion rates are reported. A postoperative air-bone gap (ABG) of <or=20 dB was obtained in 77% of the patients, 89% for the Bell (Porp) prosthesis, and 63% for the Arial (Torp) prosthesis. Overall mean pure tone averages improved 20.6 dB with ABG improvement of 19.3 dB. The overall extrusion rate was 5% (4 patients). Titanium prostheses have been easy and fast to handle and effective implants for reconstruction of the ossicular chain. We found no difference between reconstruction with or without cholesteatoma surgery during the same procedure. The combination of CWD and Torp gave significant inferior hearing thresholds as compared to Torp/CWU and Porp/CWD combinations.

Audiometry↗