Sheba Study: AI Can Spot Patients At Risk Of Pulmonary Embolism

A newly published study by Sheba Medical Center, Israel’s largest and internationally ranked hospital, shows that AI analysis of medical records as patients are admitted to the ER can accurately identify those at high risk of pulmonary embolism (PE).

A pulmonary embolism is a sudden blockage in an artery in the lung caused by a blood clot, most commonly due to a dislodged clot in the leg. They are normally diagnosed during a CT scan. 

Using machine learning, the researchers trained an algorithm to detect a pulmonary embolism before a patient was hospitalized, based on existing medical records. 

The algorithm was then used in a clinical study of more than 46,000 patients who were admitted to the ER, of whom 1,942 (4 percent) had been given a pulmonary embolism diagnosis. The study showed that the algorithm was able to accurately identify and predict which patients had a high risk of pulmonary embolism. 

The clinical study was led by Prof. Gad Segal, the head of the Sheba Education Authority, in collaboration with researchers from Ben-Gurion University of the Negev, who performed the computational development.

“Early and timely diagnosis of pulmonary embolism is challenging, yet crucial, due to the condition’s high rate of mortality and morbidity,” said Segal. 

“This study highlights the enormous potential of machine learning tools to support innovation in diagnostics. Even though the model only used data available from patients on arrival to the ER, it was still able to predict with high accuracy the likelihood of a patient developing PE, a crucial advancement for patient care and outcomes.”

The peer-reviewed research was published in the Journal of Medical Internet Research, and Sheba says the algorithm will now be integrated into its range of AI diagnostic tools for the ER.