Can ECG help doctors use artificial intelligence to detect pulmonary embolism?

Pulmonary embolism is dangerous, and pulmonary obstruction can produce blood clots

The results of the study published in the European Heart Journal-Digital Health show that the new machine learning algorithm aims to use the combination of electrocardiogram and electronic health record (EHR) data to determine whether moderate to high-risk patients really have pulmonary embolism On the one hand, it may be more effective than the currently used screening tests

The research was led by Sulaiman S.

Pulmonary embolism occurs in deep vein thrombosis, usually in the legs or arms, detaching and blocking the pulmonary artery

In order to make diagnosis easier and easier, researchers have spent more than 20 years developing advanced computer programs or algorithms designed to help doctors determine whether patients at risk actually have pulmonary embolism

In this study, the researchers found that a fusion algorithm based on EKG and EHR data may be an effective alternative method because ECG is widely used and relatively easy to manage

Researchers created and tested various algorithms based on data from 21,183 patients from the Mount Sinai Health System who showed moderate to highly suspicious signs of pulmonary embolism

The results show that the fusion model is not only better than its parent algorithm, but also better than the Wells' Criteria Revised Geneva Score and three other currently used screening tests in identifying specific pulmonary embolism cases

According to the authors, these results support the theory that ECG data can be effectively incorporated into the new pulmonary embolism screening algorithm