Improving ear health with AI

34 million children worldwide lose hearing from infections

60% of this loss is preventable

For many Indigenous children, especially those in remote communities, a simple ear infection could lead to lifelong hearing loss, impacting educational opportunities, job prospects, and quality of life. This reality affects many Indigenous children, who have the highest rate of ear disease and hearing loss globally.

About 1.5 billion people worldwide experience hearing loss, including 34 million children who often lose hearing from infections like mumps, measles, and middle ear infections. 60% of childhood hearing loss is preventable if parents are able to access ear, nose, and throat (ENT) specialists in time, but unfortunately, many Indigenous parents are not, and the consequences can be serious. Indigenous children are three times more likely to suffer from chronic suppurative otitis media, an inflammatory disease of the middle ear, and around 20-80% of these children have it. Access to ENT specialists is limited in the remote areas where Indigenous families live, delaying diagnoses and treatment. Geographic isolation often worsens health issues, making it difficult for children to receive the care they need.

Preventing hearing loss for all

Dr. Al-Rahim Habib, the lead at DrumBeat.ai, is approaching these challenges by bringing diagnostics to underserved populations using AI. DrumBeat.ai is supported by the Microsoft AI for Good Lab, and this partnership addresses ear diseases in rural and remote areas of Australia with AI and cloud computing. Dr. Habib uses DrumBeat.ai’s image-based classification algorithms to learn patterns directly from raw data, improving the diagnostic performance of non-expert practitioners. Unlike traditional methods that rely on predefined features, DrumBeat.ai’s AI-informed otoscopy—the examination of the ear using a magnification device—works amazingly well in settings with limited access to specialized healthcare.

The DrumBeat.ai application is built on a unique database of over 10,000 ear drum photos and clinical data gathered over a decade in Australia’s Northern Territory and Queensland. AI algorithms analyze ear images to detect diseases like otitis media and assign a hearing loss risk score. This predictive capability allows healthcare workers to prioritize urgent cases and refer children to ENT specialists for advanced care, or initiate treatment directly in the community.

Dr. Habib aims to integrate DrumBeat.ai into existing screening programs to enhance the efficiency, time-to-treatment, and cost-effectiveness of ear disease management in remote communities. Servicing people close to where they live empowers everyone to get involved for the health and wellness of their community.

This video contains images, voices, and names of Aboriginal and Torres Strait Islander people. We acknowledge the Kunibídji as the traditional owners of Maningrida, the land in which this was filmed. We recognize their enduring connection to the land, waters, and community, and pay respect to all traditional owners and elders, past, present, and emerging, acknowledging their rich cultural heritage and contributions.