01 Feb Artificial Intelligence for the early diagnosis of bowel cancer. What are the exciting opportunities for machine learning in health?
What are the exciting opportunities for machine learning in health? James Cameron, Head of Health and Life Sciences at Highlands and Islands Enterprise, blogs about a new project that opens up possibilities in early bowel cancer diagnosis.
Much has been said and written about the need for early diagnosis of cancer and the importance of prompt detection of symptomatic patients. This generally increases the chances for successful treatment. Great! However, does that lead to capacity challenges around clinician and other NHS staff? Of course! Is there a greater role for the rapidly-improving sophistication of artificial intelligence (AI) to ease the bottlenecks and improve outcomes? Absolutely!
One challenge is the early detection of bowel cancer, the third most commonly diagnosed cancer in both men and women in Scotland, and the second most deadly with around 1,600 dying from the disease every year.
Screening for bowel cancer in Scotland currently involves analysis of the gastrointestinal tract using traditional tube-mounted cameras (colonoscopy). Soon this may be done using images taken via ingestible cameras (colon capsules).
The traditional colonoscopy is carried out by healthcare staff, who can be prone to human errors caused by fatigue, distraction, or variable experience. The sheer volume of images and the number of patients requiring screening, places unmanageable loads on the operators which can impact on quality.
Colon capsules promise a more sophisticated solution. The images are captured in a much less invasive way and are then analysed by experts.
Inverness Campus based CorporateHealth International (CHI) – the winner of last year’s Scotland’s Life Sciences collaborative innovation award – is applying AI to aid the analysis of these images captured by colon capsules.
CHI has successfully brought together a diverse group of international collaborators to secure significant Innovate UK grant funding for a project to test and evaluate the effectiveness of this approach. CHI lead a partnership which includes Highlands and Islands Enterprise, Satellite Applications Catapult (based in Harwell), Medilogik Limited, NHS Highland, NHS Arden & Greater East Midlands commissioning support unit and the University of Barcelona.
Recent technological advances have shown that computer recognition software continuously improves (learns) as it analises more images and can spot abnormalities more effectively than human operators. The project led by CHI focuses on Video Capsule imaging of the patients internal ‘plumbing’. — Ingestible ‘capsule’ cameras — because the processes of capturing the images and their analysis are separated and therefore lend themselves to post-processing by IT systems.
For patients, it allows them to go about their daily lives rather than attending hospital or clinic – a particular benefit for Scotland’s many remote and rural communities. This saves travel time and cost and improves the experience for the patient. The system has already been trialed in Ullapool and Skye (HI-CAP). Patients who had taken bowel prep for a colonoscopy welcomed not having the long drive to hospital in Inverness without access to public toilets en-route.
Innovative projects like this helps to retain and attract young people in the Highlands and Islands by offering high quality interesting work. This project means that CHI will employ additional graduates, particularly data scientists who will develop and test the latest approaches to automated image analysis, quantify benefits to the patient, clinician and NHS– financially and clinically.
You don’t need a white coat to play a real part in helping to diagnose healthcare issues.
Sorry, the comment form is closed at this time.