Automated identification of hearing impairment based on audiogram using Pure Tone Audiometry (PTA) & Speech Recognition Threshold (SRT) using Automatic Speech Recognition (ASR) technology. Secondly, the aim
Automated Hearing Impairment Testing Device
Automated identification of hearing impairment based on audiogram using Pure Tone Audiometry (PTA) & Speech Recognition Threshold (SRT) using Automatic Speech Recognition (ASR) technology.
Secondly, the aim of this research project is to identify digitally recorded, edited & evaluated bi & mono syllabic Pashto words with the goal to identify homogeneous words that could be used to measure the SRT in native Pashto speaker of KPK, Baluchistan & Afghanistan.
The main objectives of AHIT includes:-
To help patients and audiologists who want to perform hearing impairment diagnosis by automating pure tone audiometry (for all) and Speech Recognition Thresholding (for Pushto language) system
AHIT will enable rural patients of KPK, Baluchistan & some areas of Afghanistan, to get their hearing impairment test in local language and will be served as a second opinion to the audiologist

PTA
Patient will be made heard a tone of frequency ranging 250 – 8000Hz.
On response from the patient the amp/frequency will be changed
Result will be recorded.
Result will be displayed on the screen.
SRT
Subject listens to words uttered by the machine & then repeats the same words.
Whenever the score is greater than 50%
, the level will be decremented, and whenever the score is less than 50%, the speech level will be incremented.
When the score is equal to 50% i.e three out of six words are correctly repeated, then the counter will be incremented by unity.
When the counter reaches 3, the threshold is recorded for SRT as the median of three values that incremented the counter.
Automation of entire hearing impairment testing process.
The end product of this project will help Pashto speaking community to carry out the SRT test in their native language.
The device can be programmed to perform SRT test for any language, just data set updation would be required.
The device will not only help our community in healthcare but will produce some good research work in the form of publishable research articles.
The proposed model can easily be implemented in international hospital, which will help in bringing funded projects as well as assist in international collaborations.
This project implements a novel idea that automatically identifies any hearing impairment based on automated speech recognition approach. The system uses an adaptive filter bank with weighted Mel-frequency cepstral coefficients for feature extraction. Based on the comparison the patient's speech threshold is found and displayed as a result.
The deliverables are as under:-
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Respberry pi 3 processor | Equipment | 1 | 15000 | 15000 |
| 7 Inch Capacitive Touch IPS LCD Screen For Raspberry Pi | Equipment | 1 | 12000 | 12000 |
| Signature Serious N-210 - Noise Cancelling- Water & Sweat Proof - Tang | Equipment | 1 | 5000 | 5000 |
| miscellaneous | Miscellaneous | 1 | 10000 | 10000 |
| Total in (Rs) | 42000 |
The microgrid is a small-scale, local distribution system consisting of generation...
We will construct a sensor network for detection of water contamination in agriculture, ne...
RFID, radio frequency identification is an inexpensive technology, which can be imple...
Locust swarms devastate crops and cause major agricultural damage and attendant human mise...
In the changing world, automatic facial recognition (AFR) systems have achieved numerous a...