RESUMO
PURPOSE: The anatomy and physiology of the ear is complex in nature, which makes it a challenge for audiologists to prescribe solutions for varied hearing-impaired subjects. There is a need to increase the satisfaction level of hearing-aid users by adopting better strategies that involve modern technological advancements. AIM: To design and develop a decision support Software Intelligent System (SIS) that performs audiological investigations to assess the degree of hearing loss and to suggest appropriate hearing-aid gain values. METHODS: SIS is developed based on the study conducted in the Government General Hospital, Chennai, India, between 2013 and 2015. In the study period, audiological investigations were performed on 368 subjects, using the clinical audiometer (Inventis-Piano, Italy) and the SIS. Gain suggestions were recommended for hearing-aid users (Siemens Intuis life & Intuis-SP) using standard prescriptive procedures, alterations made by the audiologists, and by the SIS. It was developed with artificial neural network-based gain predictions. RESULTS: Of the tested subjects, 256 were identified as hearing-impaired. The calculated sensitivity, specificity and accuracy of the computerised audiometer incorporated in the SIS are 93%, 85% and 90% respectively. Furthermore, 86% of the hearing-impaired subjects were satisfied during their first hearing-aid trial with the gain recommendations from SIS. CONCLUSION: The findings suggest that the proposed SIS could be used to perform audiological screening tests and to recommend appropriate hearing-aid gain values to the hearing-impaired subjects. This could eventually be helpful for audiologists in the areas where routine mass audiological screening and fast hearing-aid solution is required.
