AI algorithm differentiates normal from abnormal chest X-rays

A clinically validated and automated artificial intelligence algorithm has been shown to accurately differentiate between normal and abnormal chest X-rays.

In a study, AI chest X-ray algorithms from healthcare startup Qure.ai were trained on 1.2 million X-rays and their associated radiology reports to identify abnormalities. The deep learning system was then tested against a three-radiologist panel and a set of 2,000 independent, de-identified and anonymized X-rays.

Results of the study, recently published in a paper on Cornell University’s online research distribution site, showed that Qure.ai's qXR algorithm detects 15 chest X-ray abnormalities at near-radiologist levels with more than 90 percent accuracy.

opacity-CROP.png

“Our study shows that a deep learning algorithm trained on a large quantity of labeled data can accurately detect abnormalities on chest X-rays,” state the study’s authors. “As these systems further increase in accuracy, the feasibility of using artificial intelligence to extend the reach of chest X-ray interpretation and improve reporting efficiency will increase in tandem.”

Also See: Deep learning can be useful in screening for pulmonary diseases

In addition, the AI algorithm is seen by researchers as a valuable solution for addressing the problem of reporting backlogs and the lack of radiologists in low-resource clinical settings.

“The chest X-ray is a valuable health screening tool and a vital component of public health programs worldwide. The enormous volume produced each year creates an ever-increasing demand for radiologists,” says Shalini Govil, quality controller and senior advisor of the Columbia Asia Radiology Group. “Unfortunately, numerous chest X-rays displaying significant pathology are left neglected in piles of backlogs due to a lack of available radiologists to report them. Through semi-automation of the reporting process, AI can significantly reduce a radiologist's workload, improve report accuracy, reduce turnaround time and save lives.”

According to Prashant Warier, co-founder and CEO of Qure.ai, qXR was trained with more than 1 million curated X-rays and radiology reports, which makes it the most comprehensive and accurate chest X-ray detection algorithm currently available.

“This is an exciting time for deep learning technologies in medicine,” says Warier. “As these systems increase in accuracy, so will the viability of using deep learning to extend the reach of chest X-ray interpretation, improve reporting efficiency and save lives.”

For reprint and licensing requests for this article, click here.