You know how your doctor takes fancy pictures of your insides, like X-rays or MRIs?
These are called medical images. You might think we understand all there is to these pictures, but there’s actually a lot more data hidden in them than what our eyes can see.
Now, brainy scientists are using something called “radiomics” to get this hidden data. Radiomics is like a super-smart detective that uses AI (that’s short for artificial intelligence, by the way) to look really, really closely at medical images.
And guess what? This detective can find things that can help us understand diseases better and predict what might happen to patients.
New Study: Can AI Help Predict Heart Troubles?
In a recent study, doctors at Beth Israel Deaconess Medical Center and Lahey Hospital & Medical Center tested whether radiomics can predict heart problems in people with a common heart condition.
They used radiomics to look at MRI images of these patients’ hearts. They wanted to see if this data could help predict if these patients might have heart rhythm problems, or even sudden death.
The Heart Condition in Question
The heart condition is called hypertrophic cardiomyopathy, or HCM for short. About 1.5 million people in the U.S. have it. People with HCM have hearts with thicker than normal muscles.
This makes it harder for their hearts to pump blood to their bodies. HCM can cause symptoms like shortness of breath, irregular heart rhythm, or even sudden death.
Doctors usually use MRI to look for scar tissue in the heart of HCM patients. The presence of this scar tissue can increase the risk of sudden death.
How the Study Worked
The researchers, led by Dr. Ahmed S. Fahmy, looked at heart images from over 1,200 HCM patients. These patients were treated in Boston, Toronto, and Florence between 2003 and 2018.
The team used traditional methods to calculate each patient’s risk of sudden death within five years. These methods consider things like age, family history, the thickness of the heart wall, and how much scar tissue is present.
But then, they added radiomics into the mix. They used the MRI images to create detailed maps of the scarred heart tissue. They extracted tons of radiomic features from these maps.
They then used these features to make new mathematical models to predict the risk of sudden death.
The Results
The results showed that using radiomics significantly improved their ability to predict which patients were at high risk of sudden death.
For example, one model correctly identified 19 patients who went on to have sudden death as high risk.
Even more interestingly, the team found a specific radiomic feature that was a good predictor of sudden death risk. This feature was related to the texture of the scar tissue on the map.
This suggests that radiomics could help us find new markers, or signs, that could be used to monitor patients with HCM or other conditions that may not have clear symptoms.
What’s the Big Picture?
This study is super exciting because it shows that radiomics could be a powerful tool to predict the risk of sudden death in HCM patients. If we can predict this risk, we can potentially save lives.
More work is needed, but the early results are promising. The next step is for researchers to confirm these findings and to see if they apply to other conditions as well.
If you care about high blood pressure, please read studies about how herb medicines reduce high blood pressure, and many people with high blood pressure may take a drug that actually makes it worse.
For more information about heart health, please see recent studies that common diabetes drugs spike heart attack risk, and results showing this sleep problem linked to recurrent heart attack, stroke.
The study was published in JACC: Cardiovascular Imaging.
Follow us on Twitter for more articles about this topic.
Copyright © 2023 Knowridge Science Report. All rights reserved.
