Utilizing x-rays, radiology reports to forecast heart failure in the emergency department
Heart failure contributes to about 1 in 8 U.S. deaths each year. But experts at Massachusetts Institute of Technology have developed a new artificial intelligence platform to help combat those numbers.
MIT’s Computer Science and Artificial Intelligence Lab created its machine learning technique to analyze x-ray images and diagnose the severity of pulmonary edema, a warning sign of acute heart failure.
Using a four-point scale, with 0 being healthy and 3 very bad, the AI correctly identified levels of excess fluid in the lungs more than 50% of the time and was on-target in 90% of more severe cases
The team is working with Philips and Beth Israel Deaconess Medical Center to implement it into the Boston institution’s emergency department this fall.
“This project is meant to augment doctors’ workflow by providing additional information that can be used to inform their diagnoses as well as enable retrospective analyses,” co-lead author of a related study, Ruizhi Liao, a PhD student at MIT, said in a statement.
In that separate journal study, Liao et al. used existing x-ray cases to create new severity labels that were then used to train the machine learning system. By doing this, the model can use radiology reports to bolster its image interpretation capabilities.
The move also allows the system to explain how it arrived at its prediction, overcoming the black box problem inherent in artificial intelligence algorithms.
A paper detailing their development will be presented virtually at the International Conference on Medical Image Computing and Computer-Assisted Intervention on Oct. 5.
Matt joined Chicago’s TriMed team in 2018 covering all areas of health imaging after two years reporting on the hospital field. He holds a bachelor’s in English from UIC, and enjoys a good cup of coffee and an interesting documentary.