Heart Attack: Better prediction with New Artificial Intelligence enabled tool

Last updated on August 3rd, 2019 at 06:33 pm

Soon heart attacks and other cardiac issues will be more accurately predicted using a tool developed by researchers that uses AI (Artificial Intelligence) and Machine Learning (ML) algorithm.

According to recent research published on 25 June 2019 on RADIOLOGY, the researcher found that Machine Learning methods better discriminated patients who subsequently experienced a cardiac disorder from those who did not.

Research methodology

Presently, the data extracted from Coronary CT angiography is studied on basis of Coronary Artery Disease Reporting and Data System (CAD-RADS) score to predict heart attacks.

But, scientist wanted to prove that coronary CT angiography vessel score generated by Machine Learning would be superior to CAD-RADS and other conventional scores in discriminating patients who experienced a subsequent adverse event from those who did not.

The research was carried out at US, there were a total number of 6892 patients were included in the research. They were selected by reviewing there CT angiography from February 2004 through November 2009 complied with Health Insurance Portability and Accountability Act guidelines.

All these 6892 patient has following indications ranging from:

• Atypical chest pain
• Indeterminate stress test,
• Multiple risk factors,
• Strong family history,
• And strong personal concern for coronary artery disease.

Images from CT angiography of all the patients were studied and interpreted by group of five highly qualified and experienced radiologist. They interpreted the data and computed five conventional vessel (blood vessel) scores.

1. CAD-RADS
2. Segment stenosis score.
3. Segment involvement score.
4. CT Leaman score.
5. Segment plaque burden score.

However, researchers also studied the data derived from Machine Learning vessel score and compared it with five of the above conventional vessel scores.

Results

The results confirm that the machine learning models did better discriminate these patients for the prediction of coronary artery deaths on the basis of coronary CT angiography.

Keep reading: CT scan vs MRI: Purpose is same, What’s the difference?

Resource:

1. Scoring of Coronary Artery Disease Characteristics on Coronary CT Angiograms by Using Machine Learning (Published Online:June 25 2019): https://pubs.rsna.org/doi/10.1148/radiol.2019182061