Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

AI can detect pneumonia key findings in chest X-ray scans in 10 seconds

Researchers from Intermountain Healthcare and Stanford University have created a new artificial intelligence (AI) system that would be able to reduce chest X-rays scanning times considerably.

The new study analysed the performance of the new program and concluded it was able to accurately identify key findings in chest X-rays of patients in the emergency department suspected of having pneumonia in just 10 seconds. For context, traditional AI scanning systems would take around 20 minutes to complete this task.

The researchers said this new scanning speed may enable physicians to accurately confirm a pneumonia diagnosis way before traditional clinical practice, thus enabling to start treatment for the condition sooner, which is very important for severely ill patients who’re suffering from pneumonia.

The findings from the collaborative study were showcased at the European Respiratory Society’s International Congress in Madrid, Spain, on Monday.

The new AI system is called CheXpert and was developed by the Stanford Machine Learning Group to review X-ray images taken at several emergency departments at Intermountain hospitals throughout Utah. It was trained using 188,000 chest imaging studies.

“CheXpert is going to be faster and as accurate as radiologists viewing the studies,” said Nathan C. Dean, MD, principal investigator of the study, and section chief of pulmonary and critical care medicine at the Intermountain Medical Center.

“It’s an exciting new way of thinking about diagnosing and treating patients to provide the very best care possible.”

Dean added that their research also revealed that often the accuracy of traditional medical scanning system was hindered by human error, and that AI could reduce that error substantially.

“A 2013 study published in JAMA Internal Medicine found that 59 per cent of errors made by ePNA were due to NLP processing of radiologist reports, so we’re eager to replace it with a better, faster system,” Dean explained.

The next step, he concluded, is for the research model to be deployed in actual emergency departments, which he believes will happen in certain Intermountain Healthcare hospitals this fall.

Image via Freepik on flaticon.


Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter