The researchers reviewed data from an estimated 500,000 anonymized adult chest X-rays to develop an AI system that recognizes abnormalities in the images. The AI system not only detects abnormalities from X-ray images, but also assesses their urgency, allowing it to prioritize how quickly the exams should be reviewed by a radiologist.
To validate the AI system, the researchers tested it with historical radiological exams.
The AI system successfully distinguished normal and abnormal chest X-rays, according to study results published in the journal Radiology. Due to the speed at which the AI system assessed X-rays, the researchers suggested abnormal images with critical findings could have reached expert radiologists for opinions sooner than typical practice, cutting the average delay from 11 days to less than three days.
“Artificial intelligence led reporting of imaging could be a valuable tool to improve department workflow and workforce efficiency,” said Giovanni Montana, PhD, data science chair of Warwick Manufacturing Group at the University of Warwick and lead author on the study.
“The results of this research shows that alternative models of care, such as computer vision algorithms, could be used to greatly reduce delays in the process of identifying and acting on abnormal X-rays,” he continued.
More articles on clinical leadership and infection control:
Researchers develop AI tool for UTI detection
9 states report high flu activity: 5 things to know
Physician’s encouraging words more effective than many may realize, study finds
