How AI is Revolutionizing Early Diagnosis: A Case of Osteoporosis Detection and the Future of Healthcare

Will Studholme, 58, was admitted to an NHS hospital in Oxford in 2023 for treatment of severe stomach issues. What he didn't expect was to depart with a diagnosis of osteoporosis—a disease commonly associated with age that weakens bones and raises the risk of fractures. Although Mr. Studholme initially sought care for food poisoning, an abdominal CT scan revealed something unexpected: a collapsed vertebra in his spine, an early indicator of osteoporosis.

It was a sophisticated AI system that powered it that subsequently analyzed the scan, highlighting the potential problem. "I feel very lucky," says Mr. Studholme. "I don't think this would have been picked up without the AI technology." Further tests followed and treatment with an osteoporosis drug that is bound to improve his bone density and prevent fractures in the future.

Role of AI in Opportunistic Imaging:

It so happens that radiologists are taught to look for specific problem areas in the scan. The idea of AI finding other, undiagnosed health problems is a promising new development in medical care. This is known as opportunistic imaging: scanning for signs of common, chronic diseases not related to the reason the scan was done. The clinical use of AI for opportunistic screening or opportunistic imaging is just beginning," says Prof Perry Pickhardt, professor of radiology at the University of Wisconsin-Madison.

Opportunistic imaging uses scans already completed—as part of evaluation screens for suspected cancer or to look at symptoms that may seem linked to a related problem, such as abdominal pain—to check for other things, like osteoporosis, heart disease, fatty liver disease, and even diabetes. The beauty of the method is that it can identify problems earlier in life, even before symptoms appear when the treatment is far more effective. "We can avoid a lot of the lack of prevention that we have missed out on previously," says Prof. Pickhardt. Many chronic diseases often go undetected by routine physical exams or blood tests. This is where this particular approach is also equally valuable.

The Future of AI and Healthcare: Reducing Bias:

Miriam Bredella, a radiologist at NYU Langone who is also working on AI algorithms, points out another significant advantage of AI technology: the reduction of bias. For instance, osteoporosis is often considered to be a disease of elderly, thin white women, and doctors tend to miss other populations, such as younger men. AI does not carry such biases, ensuring that a wider range of patients receive timely diagnoses. Mr. Studholme is an example of the same point-he was younger, male, and had no previous history of fractures, making it impossible to identify his osteoporosis without AI intervention.

The Rise of AI in Medical Diagnosis:

AI is not limited to osteoporosis detection. Other health issues are being explored, such as cardiovascular disease, liver problems, and age-associated muscle loss. While most AI research has concentrated on CT scanning of the stomach or chest, attempts are now underway to apply AI to other imaging modalities such as X-rays of the chest and mammograms.

The AI systems that have led to these achievements are trained on massive scan datasets, guaranteeing that the technology performs effectively across a wide range of ethnic groups.

This is important because, for AI to be successful on a global scale, it must be trained to recognize health issues in a variety of populations. There's always a human element, however: when AI detects something unusual, the results are reviewed by radiologists to confirm accuracy before they are shared with doctors

Nanox.AI: The Future of AI in Medicine

The AI technology used for Mr. Studholme's scan comes from Nanox.AI, an Israeli company focused on opportunistic screening products that can identify osteoporosis, heart disease, and fatty liver disease from routine CT scans. Oxford NHS hospitals began trialing Nanox.AI’s osteoporosis-focused product as early as 2018, and by 2020, it was officially rolled out. Early results showed a significant increase—up to six times the NHS average—in the detection of vertebral fractures, leading to earlier diagnoses of osteoporosis and, consequently, earlier treatment.

According to Professor Kassim Javaid, who initiated the use of Nanox.AI's algorithm at Oxford, such results may contribute to better diagnosis and treatment of osteoporosis in the NHS. "We want to build the evidence to use it across the NHS," he says. Other trials are currently being conducted at Cambridge, Cardiff, Nottingham, and Southampton hospitals.

The Challenges of AI in Healthcare

While the potential for AI in early diagnosis is great, a few challenges come to mind. According to Prof. Sebastien Ourselin, an expert in healthcare engineering at Kings College London, "This is increasing the demand on the healthcare system, not reducing it," he says. The flagging of more patients by AI for further testing might result in a flood of diagnostic procedures, especially if the AI systems are over-sensitive.

 Furthermore, new services would be required to handle the increased number of people diagnosed with conditions such as osteoporosis, heart disease, and fatty liver disease. Solutions are already being developed in Oxford to address this challenge. The following example is a service for patients suffering from osteoporosis fractures whereby the patient would be referred to a fracture prevention service managed by nurses, releasing doctors to be able to engage in other functions.

Prof. Javaid states, "The AI does force you to change your pathway, so the whole health service can absorb it."

Benefits of Early Diagnosis in the Long Run:

Despite these challenges, using AI for early detection will eventually reduce costs and improve outcomes. Early intervention in conditions like osteoporosis could prevent costly fractures, which are one of the top causes of hospital admissions.

"Fracture is one of the top reasons people end up in hospital," says Prof. Javaid.

By identifying more cases early on, the NHS could reduce the long-term burden on healthcare services.

Mr. Studholme personally experienced the impacts of AI technology, which changed his life. He is grateful to have received treatment before his bones became brittle, as his mother suffered from multiple fractures from osteoporosis.

"I feel quite privileged I can do something before my bones turn into chalk," he says.

AI is undoubtedly going to change the face of medicine in diagnostics; as its role grows, it will revolutionize how we spot and prevent diseases, leading to better health benefits for individuals and communities.

Source: BBC