Our client required development of an application to form part of their remote hearing test kit. Utilising a machine-learning algorithm to identify the level of hearing loss in a subject, the kit enables people to administer their own hearing tests with clinical accuracy, for use in occupational health settings. Alongside perfectly consistent functionality and a considered UX design, the application required full compliance with medical and security standards.

What we did

Settling on creating a web application that prioritises versatility, our experienced development team selected Next.js and .NET frameworks for this unique build. Within the first month of development, an admin panel had already been created from the ground up.

The machine-learning algorithms for testing hearing health were developed by leading audiologists and provided to our team as a black box program. We then wrote the API logic that wrapped this code, ensuring everything interfaced perfectly with both the cloud-based data-collection system and the custom, user-friendly UI.

A calibration function was a necessary addition, so as to eliminate inconsistencies in the tests due to hardware discrepancies. The calibration system was provided by audiology company Guymark and implemented by our development team. Given the confidentiality of the data this application is designed to collect, we also arranged penetration testing with Astra Security and followed OWASP best practices throughout.

We carried out end-to-end testing in order to assess the entire application flow, working closely with the client and their legal counsel to provide regular updates, discuss features and ensure the software stood up to the rigorous standards expected from medical equipment. Accordingly, the test kit is certified as a Class I medical device, as described in the UK MDR 2002.