AI in Healthcare: how AI is revolutionising the field of medical imaging

AI in Healthcare: how AI is revolutionising the field of medical imaging

This installment highlights promising advances in the use of AI in the field of medical imaging.

Why do we need AI in medical imaging?

Medical imaging technologies allow healthcare professionals to diagnose, treat and manage disease. Examples of such technologies include X-Rays, MRI scans, ultrasound scans, fluoroscopy, and mammograms. The importance of these diagnostic tools in delivering beneficial outcomes for patients means clinical radiologists undergo rigorous specialist training – typically an additional 7 years upon completing medical school.
Despite receiving such extensive training, human error can lead to clinicians providing patients with inaccurate diagnoses. Furthermore, for life-threatening conditions such as cancer, images are analysed by two radiologists, with the second radiologist blind to the clinical decision of the first. Overall, these procedures contribute to prolonged wait times for patients and may lead to conflicting diagnoses.
In the interests of increased precision, procedural efficiency, and improved patient outcomes, medical imaging is certainly one aspect of healthcare that will continue to benefit from the ongoing AI revolution.

What is the state of the art in this area?

In 2019, the UK government announced they would be investing £250 million to create a new centre for the integration of AI into UK healthcare services. One of the key directives of this centre, now known as the “NHS AI Lab”, has been to evaluate new AI-based approaches to improve screening methods for various diseases. Areas where there are already highly promising results include mammography and CT scanning.
Kherion Medical Technologies have developed a diagnostic tool called Mammography Intelligent Assessment (MIA®), which utilises convolutional neural networks trained using over 3 million mammogram images from various sources. The MIA® suite of AI models reduces the need for unnecessary recalls due to low quality images, flags suspicious cases, and improves mammogram scheduling. Additionally, as every mammogram performed by the NHS is traditionally checked by two radiologists, the implementation of MIA® in the NHS has the potential to minimise delays in the existing breast cancer screening system. Overall, MIA® significantly reduces the amount of time patients are waiting for their results, which reduces the stress and anxiety associated with awaiting diagnoses – in particular as most women who receive mammograms will receive a negative test result.
AI Skunkworks is an NHS AI Lab programme focussed on supporting the early adoption of AI approaches. In 2021, AI Skunkworks collaborated with George Eliot Hospital and Roke (a UK-based research & development company), to evaluate the potential of AI for improving the early detection of abnormal growths in computerised tomography (CT) scanning. Results showed that various machine learning algorithms were effective in differentiating between tissue types, detecting small new growths, and aligning scans taken at different time points. Whilst their preliminary results are promising, the consortium is currently seeking further funding to test whether these approaches can be safely and effectively integrated into the modern-day radiologist’s toolkit.

What does the future look like?

Although there has been increased research activity into AI-based medical imaging since the early-mid 2010s, many of these approaches need to be improved further before they are used in real-world clinical settings. Nonetheless, we continue to see growing numbers of patents filed in this area – in particular, in the United States and China.
Perhaps it is misunderstanding of the European Patent Office’s (EPO) stance on the patentability of software which has led to the comparatively low number of European filings in this area. Whilst computer programs and the like are excluded from patentability in Europe per se, decades-worth of case law at the EPO makes it clear that computer-implemented inventions (and inventions related to AI) are indeed patentable in Europe as long as they solve a technical problem. This is certainly the case for improvements in medical imaging and healthcare more generally.

Closing remarks

AI is becoming more embedded into every aspect of everyday life and the healthcare sector is no exception. The team at Kilburn & Strode is excited to be supporting innovators in this area, who are pioneering the AI revolution in healthcare. Stay tuned for our next installment, where we will be discussing applications of AI in clinical trials.
For further information or advice related to patenting inventions in this area, please do not hesitate to contact Dan Olatokun or your usual Kilburn & Strode advisor.
If you enjoy our series and wish to discuss any of the related content or are interested in protecting your own innovation in these areas, do reach out to our Bioinformatics and Digital Health team

Second installment: Coming soon!

Let us keep you up to date. If you’d like to receive communications from us, ranging from breaking news to technical updates, thought leadership to event invitations, please let us know.

Connect with us