On March 30, 2022 – Radiology researchers found that AI is a useful fracture detection tool that could help busy emergency department clinicians stay on top of their game during rush hour.
Fractures can be missed or misdiagnosed on X-rays, which can have serious consequences for the patient. The problem is exacerbated further by the inability to obtain expert advice in a timely manner, as imaging volume growth continues to outpace recruitment of radiologists.
Artificial Intelligence (AI) may be able to assist radiologists in the diagnosis of fractures by acting as an assistant.
To better understand the potential of AI in the fracture setting, researchers in England reviewed 42 existing studies comparing AI’s diagnostic performance with that of clinicians in the detection of fractures. When it came to identifying fractures, 37 of the 42 studies used X-rays, while five used CT.
Also read: Investment of *350 crore by Sundram Fasteners in advanced technology
The researchers found no differences in performance between clinicians and AI that could be attributed to chance. Fractures were detected by AI with a 91% to 92% sensitivity rate.
According to the study’s lead author, Rachel Kuo, M.B., B.Chir., from the Botnar Research Centre in Oxford’s Nuffield Department of Orthopaedics and Rheumatology, AI performed “with a high degree of accuracy that was comparable to clinician performance,” she said in a news release. We found this to be true when AI was validated using separate external datasets, suggesting that the findings could be applied to a broader population.
According to Dr. Kuo’s findings in the study, artificial intelligence (AI) may have numerous educational and clinical applications in fracture detection. It could reduce the rate of early misdiagnosis in difficult emergency situations, such as patients who have multiple fractures. It has the potential to serve as a teaching tool for aspiring clinicians.
“It could also be helpful as a “second reader,” providing clinicians with either reassurance that they have made the correct diagnosis or prompting them to take another look at the imaging before treating patients,” said Dr. Kuo.
Dr. Kuo cautioned that AI fracture detection research is still in its very early, preclinical stages and should not be taken too seriously just yet. There was only one study where an AI was evaluated in a clinical setting in a prospective study, and only a small percentage of the studies examined the performance of clinicians with AI assistance.
It’s still important for clinicians to use their own judgement, said Dr. Kuo. Artificial intelligence (AI) is not perfect and is prone to bias and error.
Also read: IBM Edge Application Manager is now accessible to smart devices with mimik Technology edgeEngine