Getty ImagesThis is the sixth feature in a six-part series examining how AI will transform medical research and treatment.
Will Studholme, 58, never expected to be diagnosed with osteoporosis when he was rushed to Oxford NHS Hospital in an accident with gastrointestinal symptoms in 2023.
This disease is strongly associated with aging, causing bones to become weaker and more brittle, increasing the risk of fractures.
Mr Studholm was found to have suffered from severe food poisoning and underwent an abdominal CT scan during the early stages of investigating his illness.
The scan was then run by artificial intelligence (AI) technology and identified collapsed vertebrae in Mr Studholm’s spine, a common early indicator of osteoporosis.
After further testing, Studholm not only had a diagnosis, but also a simple treatment: an annual infusion of osteoporosis medication that is expected to improve bone density.
“I feel very lucky,” Studholm says. “I don’t think this would have been picked up without AI technology.”
Will StudholmIt is not unheard of for radiologists to discover something incidental in a patient’s imaging that is not what they were originally looking for (an undetected tumor, a concern about a particular tissue or organ).
But behind the scenes, we apply AI to systematically comb through scans and identify the early stages of common preventable chronic diseases that may be occurring, regardless of why the scan was ordered in the first place. Identifying symptoms automatically is new.
The clinical use of AI, known as opportunistic screening or opportunistic imaging, is “in its infancy,” said Perry, a professor of radiology and medical physics at UW-Madison and one of the algorithm’s developers. Pickhart points out.
This is considered opportunistic as it takes advantage of imaging already being done for other clinical purposes, such as suspected cancer, chest infections, appendicitis, and abdominal pain.
It may be possible to detect previously undiagnosed diseases at an early stage, before symptoms appear, making it easier to treat and prevent disease progression. “We’ll be able to avoid a lot of the lack of precautions that we’ve missed in the past,” Professor Pickhart said.
These diseases often go undetected during routine physical exams and blood tests, he added.
Getty ImagesCT scans have a lot of data related to body tissues and organs that we don’t actually use, says Miriam, a radiologist at NYU Langone who also develops algorithms in this field.・Mr. Bredera points out.
And in theory, that analysis could be done by radiologists taking measurements without AI, but that would take time.
She points out that the technology also has benefits in terms of reducing bias.
For example, a disease like osteoporosis is thought to primarily affect thin, older, white women, so doctors don’t necessarily think to look beyond that population.
Opportunistic imaging, on the other hand, does not discriminate as such.
Mr Studholm’s case is a good example. He was relatively young for an osteoporotic patient, male, and had no history of fractures, so it is likely that he would not have been diagnosed without the AI.
In addition to osteoporosis, AI is being trained to opportunistically identify heart disease, fatty liver disease, age-related muscle loss, and diabetes.
The main focus is on CT scans, such as those of the abdomen and chest, but efforts are also being made to gather information from other types of images, such as chest X-rays and mammography.
The algorithm is trained on thousands of tagged past scans, and it is important that the training data includes scans from a wide range of ethnic groups if the technology is to be introduced to a diverse population. , experts emphasize.
A certain amount of human testing will then be conducted. If the AI finds something suspicious, it will be sent to a radiologist for confirmation before being reported to a doctor.
The AI technology used to examine Studholm’s scans comes from Israeli company Nanox.AI, one of the few companies working on AI for opportunistic screening. Far more companies are focused on using AI to aid in accurate and rapid diagnosis. The specific conditions under which the scan will actually occur.
Nanox.AI offers three opportunistic screening products that aim to help identify osteoporosis, heart disease, and fatty liver disease, respectively, from routine CT scans.
Oxford NHS Hospitals began piloting Nanox.AI’s osteoporosis-focused product in 2018, before formally rolling it out in 2020.
Findings from an Oxford hospital show that the number of patients identified with vertebral fractures is up to six times higher than the NHS average, meaning patients can now be tested for osteoporosis and start treatment to combat it. Professor Qasim Javaid said. He is a researcher in osteoporosis and rare bone diseases at the University of Oxford and has been at the forefront of introducing this algorithm.
Further trials of the algorithm are currently underway in hospitals in Cambridge, Cardiff, Nottingham and Southampton. “We want to build the evidence for its use across the NHS,” Professor Javaid says.
But while this technology has the potential to benefit individuals, there are also broader implications that need to be considered. Sebastian Ursulin, head of the AI Center, said:
A big issue that needs to be balanced, he points out, is the potential increase in patient numbers through the use of technology. “This is increasing demand on the health care system, not decreasing it,” he says.
First, people flagged as possibly having the disease through opportunistic screening are likely to require further confirmatory testing, which takes resources. Also, if the AI is inaccurate or too sensitive, it can lead to a large number of unnecessary tests.
And we need to provide services for additional people who end up being diagnosed.
Professor Javaid acknowledges that the extra load is a challenge with the technology, but there are solutions.
Patients with confirmed fractures in Oxford are referred for follow-up to the fracture prevention service, which is primarily staffed by nurses, to avoid overburdening doctors. “AI does force rerouting,” he says.
And in the long term, Professor Jaid believes it will save the NHS money by ensuring more people with early-stage osteoporosis are identified and receive the preventive treatment they need. “Fractures are one of the main reasons people end up in the hospital,” he says.
Studholm has seen firsthand the devastation of osteoporosis, which led to his mother breaking both hips. Previously, it was thought to be just a symptom of the elderly and nothing could be done about it, he says. “I’m so honored to be able to do something before the bones turn to chalk,” he says.
