The NHS is rolling out AI cancer diagnosis tools nationally, demonstrating the technology’s ability to accelerate cancer diagnosis.
In a recent announcement, the UK’s National Health Service (NHS) will utilize AI to analyze X-rays as part of a £21 million initiative to accelerate cancer diagnosis and reduce waiting times.
AI-powered medical diagnostic tools have gone from strength to strength. Throughout the 2010s, AI computer vision (CV) models demonstrated their ability to interpret medical scans similarly to humans. In 2018, researchers from DeepMind, UCL, and Moorfields Eye Hospital developed a breakthrough AI model to diagnose 50 eye diseases as accurately as a doctor.
From there, AI-supported medical imaging and diagnostics have evolved rapidly, with numerous models demonstrating better accuracy than human specialists.
Unlike often overworked radiologists, AIs work tirelessly, relieving the burden on human teams and enabling them to deliver enhanced care and treatment.
On Friday, UK Health Secretary Steve Barclay announced plans for deploying AI-assisted medical diagnostics for lung cancer before winter this year. Over 20 NHS sites have already started implementing this technology, with preliminary results indicating it may be 40 times more precise than traditional techniques, yielding results in under 30 seconds. Over 600,000 X-rays are conducted monthly across England.
Furthermore, hospitals will start implementing AI technology for stroke diagnosis, leading to patients receiving treatment about an hour earlier on average and tripling the recovery rate.
Barclay said, “Artificial intelligence is already transforming the way we deliver healthcare, and AI tools are already making a significant impact across the NHS in diagnosing conditions earlier, meaning people can be treated more quickly.”
Health officials hope AI and other technologies will help the UK healthcare system alleviate backlogs and staff pressure.
NHS national medical director, Professor Stephen Powis, added, “The NHS is already harnessing the benefits of AI across the country in helping to catch and treat major diseases earlier, as well as better managing waiting lists so patients can be seen quicker.”
How AI is used in medical diagnostics
AI Convolutional Neural Networks (CNNs), a form of deep learning model, are transforming medical diagnostics. Interpreting scans is no easy task, and studies observe error rates of around 3 to 5% or higher when radiologists are fatigued or overworked.
CNN models work similarly to neurons in the human brain, discerning fundamental features from scans, such as edges or textures, and recognizing more complex patterns, like the distinct shape and size of lung nodules that might indicate cancer.
One of the first major breakthroughs in AI-supported medical imaging was the LUNA16 (LUng Nodule Analysis 2016) Competition, where participants developed methods of identifying and diagnosing lung cancers from X-ray and CT images.
Many of the algorithms that emerged from this competition matched – or exceeded – the diagnostic capabilities of trained radiologists.
Soon after, Google’s DeepMind emerged as a chief innovator in AI medical imaging, applying the technology in fields like eye disease detection and breast cancer screening. DeepMind’s Alphafold uses AI-powered 3D modeling to further our understanding of how proteins work, accelerating cancer research.
Now, the NHS hopes to demonstrate how cutting-edge AI medical technology can be rolled out across entire healthcare systems.
