0
0
Edinburgh, Scotland—In a promising leap for global health equity, a Heriot-Watt University researcher is developing artificial intelligence technology that could transform skin cancer diagnosis for patients living far from medical centers.
For the past two years, Ph.D. candidate Tess Watt has led the charge to design AI-driven diagnostic tools targeting communities with limited access to dermatological care. Working from the university’s School of Mathematical and Computer Sciences, Watt aims to empower individuals in regions as diverse as rural Scotland and West Africa to detect skin cancer and other conditions early—even in the absence of internet connectivity.
“Health care from home is a crucial topic right now, especially with growing GP wait times,” says Watt. “If we can empower people to monitor skin conditions from their own homes using AI, we can dramatically reduce delays in diagnosis”.
How the System Works
The system, which is reportedly the first of its kind to blend AI diagnosis with a mission to serve remote communities, uses affordable Raspberry Pi computers. Patients photograph their skin concern using a camera attached to the device. AI algorithms, pre-loaded with vast datasets of skin images, analyze the photo in real time—flagging potential issues for medical review, all without needing an internet connection.
This autonomous setup offers up to 85% diagnostic accuracy and aims to improve further as additional datasets and advanced machine learning models are integrated. After analysis, results are shared with local GP services to kickstart any necessary treatment.
Advancing Toward Real-World Impact
Watt’s prototype has already been demonstrated in Heriot-Watt’s advanced health and care technologies suite. While clinical testing is still pending, talks are underway with NHS Scotland to initiate the ethical approval process, with hopes for a pilot project in the coming year or two.
Her vision is bold—rolling the system out through Scotland’s remote areas before expanding internationally to regions where travel to a doctor is difficult or impossible. The technology could also provide lifelines for the housebound and infirm, with family help facilitating the diagnostic process.
Designing for Resilience
A key feature, emphasized by academic supervisor Dr. Christos Chrysoulas, is the device’s ability to operate independently of the internet. “E-health devices must be engineered to operate independently of external connectivity to ensure continuity of patient service and safety,” notes Chrysoulas. “In the event of a network or cloud service failure, such devices must fail safely and maintain all essential clinical operations without functional degradation”.
This resilience is vital for ensuring ongoing patient care in places where connectivity is unreliable or non-existent.
Looking Ahead
Watt, whose background includes accessible translation technologies and research in Tiny Machine Learning, sees her work as essential amid surging health service backlogs and resource constraints. “By the time I finish my Ph.D., three years from now, I’d love to see something well into the pipeline that’s on its way to real-world use,” she shares.
The project aligns with Heriot-Watt’s broader commitment to addressing critical health challenges through its Global Research Institute in Health and Care Technologies, which also tackles issues from antimicrobial resistance to emerging epidemics.
Disclaimer: The technology described is in the early stages of development and has yet to be validated in clinical settings. Diagnostic accuracy, regulatory approval, and real-world efficacy remain to be fully established. Patients should continue to seek guidance from qualified health professionals regarding any concerns about skin conditions or cancer at this time.
