This feature is the sixth installment of a series exploring the impact of AI on medical research and treatments.
When 58-year-old Will Studholme visited an NHS hospital in Oxford in 2023 due to gastrointestinal issues, he received an unexpected diagnosis of osteoporosis. This condition, commonly linked with aging, results in fragile and weak bones, making fractures more likely.
Although Mr. Studholme actually had severe food poisoning, during the initial evaluation, he underwent an abdominal CT scan. By utilizing artificial intelligence (AI) technology, the scan identified a collapsed vertebra in his spine, a key early indicator of osteoporosis.
Following further tests, Mr. Studholme not only received his diagnosis but also a straightforward treatment plan: an annual infusion of an osteoporosis medication anticipated to enhance his bone density.
Expressing his gratitude, Mr. Studholme stated, "I feel very lucky. I don't think this would have been picked up without the AI technology."
While radiologists may incidentally notice abnormalities in imaging, AI technology presents a novel approach by systematically analyzing scans for early signs of common preventable chronic diseases. This method, known as opportunistic screening or imaging, is in its initial stages of deployment.
Perry Pickhardt, a professor specializing in radiology and medical physics at the University of Wisconsin-Madison, acknowledges the emergence of this approach, stating that it leverages existing imaging data for various clinical purposes.
By detecting potential diseases in their early stages before symptoms manifest, opportunistic imaging can facilitate prompt treatment or prevention. This approach aims to address gaps in traditional methods such as regular physical exams or blood tests, which may not always detect these conditions.
Miriam Bredella, a radiologist at NYU Langone, highlights the unused potential in CT scans and underscores the efficiency of AI algorithms in analyzing vast amounts of data rapidly. Notably, AI technology can help mitigate biases in diagnosis, ensuring a more inclusive approach.
The remarkable case of Mr. Studholme exemplifies how AI technology can identify diseases like osteoporosis in unexpected demographics. Beyond osteoporosis, AI is also being trained to detect heart disease, fatty liver disease, age-related muscle loss, and diabetes through opportunistic screening.
Although current focus lies on CT scans, efforts are underway to expand the application to other imaging modalities such as chest X-rays and mammograms. Training algorithms on diverse data sets is crucial for deploying this technology effectively across different population groups.
Ensuring a human oversight component is essential in the process, as flagged findings by AI require validation by radiologists before clinical reporting. Companies like Nanox.AI are pioneering the development of AI for opportunistic screening, aiming to enhance diagnostic capabilities in healthcare settings.
Oxford NHS hospitals' successful trials with Nanox.AI's osteoporosis screening product underscore its potential for early disease detection and management. Collaborative efforts across various hospitals aim to establish a robust framework for deploying AI technology in healthcare systems.
While the benefits of AI in healthcare are evident, experts like Sebastien Ourselin from Kings College London advise a balanced approach due to potential challenges such as increased patient volumes and resource demands. Adapting care pathways to accommodate the influx of patients flagged by AI is crucial for optimizing healthcare delivery.
For individuals like Mr. Studholme, early detection and intervention offered by AI technology present a proactive approach to managing conditions like osteoporosis. By identifying and treating these conditions in their initial stages, healthcare systems can potentially reduce long-term costs and improve patient outcomes.