Global access to digits-in-noise hearing tests

 This project focuses on advancing hearing test technology by automating the development of the digit-in-noise (DIN) test through artificial intelligence. The public-private partnership with Cochlear combines the expertise of various stakeholders to streamline the creation of these tests, which are used for screening and as a diagnostic tool for those with cochlear implants (CI). 

Hearing loss affects millions globally, with over 466 million people worldwide suffering from disabling hearing loss, according to the World Health Organization. As populations age, this number is expected to rise to 900 million by 2050. Early diagnosis and effective testing are critical in preventing further hearing damage. However, developing hearing tests that are language-specific can be costly and time-consuming, limiting accessibility, especially in low- and middle-income countries. By reducing the cost and time required to develop these tests, this project addresses an important global health challenge. 

The project uses artificial intelligence to automate the development of DIN tests, making the process faster and more cost-effective. It investigated the suitability of synthetic speech in five languages, demonstrating that its accuracy was comparable to natural speech across the board. Additionally, automatic speech recognition systems (ASRs) were tested to standardize the intelligibility of synthetic digits, with one particular system performing very well. This innovative approach ensures that the tests can be rapidly deployed in multiple languages. 

The project successfully created fully automated DIN tests in both English and Dutch, which were compared with standard reference tests. Both AI-driven tests demonstrated high classification accuracy for participants with hearing loss. Ultimately, the Aladdin project has made the development of DIN tests more affordable and accessible, especially for low- and middle-income countries, by significantly lowering production costs.