At the annual meeting of the Radiological Society of North America (RSNA), researchers revealed a deep learning model that uses just one chest X-ray to predict the 10-year risk of heart attack or stroke due to atherosclerotic cardiovascular disease.
What is Deep Learning?
Deep learning is a subset of artificial intelligence (AI) that can be programmed to recognize patterns associated with diseases in X-ray images.
Dr. Jakob Weiss, the study’s lead author and a radiologist affiliated with Massachusetts General Hospital and the AI in Medicine program at the Brigham and Women’s Hospital in Boston, emphasized the potential of this model in providing population-based screening for cardiovascular disease risk using available chest X-ray images.
He mentioned that this could help in identifying patients who are not on statin medication but might benefit from it.
To decide who should be on statin medication for primary prevention, it’s recommended to estimate the 10-year risk of significant adverse cardiovascular disease events.
This estimation is based on the atherosclerotic cardiovascular disease (ASCVD) risk score, which takes into account various factors, including age, sex, race, blood pressure, and more.
Patients with a risk of 7.5% or higher over the next decade are typically recommended for statin treatment.
The Problem
Often, the variables needed to calculate the ASCVD risk are not readily available, which highlights the need for broad-based screening methods.
With the widespread availability of chest X-rays, the new deep learning approach could fill this gap.
The research team trained their model, named CXR-CVD risk, using 147,497 chest X-rays from 40,643 participants in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial. This trial was sponsored by the National Cancer Institute.
The model was tested on a separate group of 11,430 outpatients, of which 1,096 (or 9.6%) experienced a major cardiac event over an average follow-up period of 10.3 years.
The CXR-CVD risk deep learning model showed a notable correlation between its predicted risks and the observed major cardiac events.
When compared with the traditional clinical standard for deciding statin eligibility, which could only be calculated for 21% of the patients due to incomplete data, the CXR-CVD risk model showcased similar performance and even added value.
Conclusion and Future Steps
Dr. Weiss highlighted the model’s simplicity, emphasizing that a single X-ray, a procedure done millions of times daily worldwide, can predict cardiovascular events comparably to established clinical standards.
However, further research, including controlled, randomized trials, is essential to validate this deep learning model. Ultimately, this model could serve as a decision-support tool for physicians.
The broader message Dr. Weiss conveyed was that a chest X-ray is not just a tool for diagnosing lung-related issues; it has the potential to offer both diagnostic and prognostic information that can benefit both doctors and patients.
If you care about stroke, please read studies about how to eat to prevent stroke, and scientists find a breakfast linked to better blood vessel health.
For more information about health, please see recent studies about how Mediterranean diet could protect your brain health, and wild blueberries can benefit your heart and brain.
Follow us on Twitter for more articles about this topic.
Copyright © 2023 Knowridge Science Report. All rights reserved.
