- An estimated 15 million people worldwide suffer a stroke each year.
- A person may have a better outcome after a stroke if treated quickly.
- In a new study, researchers used artificial intelligence and a single chest X-ray to predict a person’s 10-year risk of dying from a stroke or heart attack.
Every year, around 15 million people worldwide suffer a stroke, which causes around 5 million deaths and permanently disables another 5 million.
A person’s outcome after a stroke depends on how quickly they get treatment once symptoms start.
Recently, researchers at Massachusetts General Hospital developed a deep learning model that uses artificial intelligence (AI) and a single chest X-ray to predict a person’s 10-year risk of dying from a stroke or heart attack.
The results of the study were presented at the annual meeting of the Radiological Society of North America (RSNA) in Chicago at the end of November.
What you know about a stroke
Should Stroke is a type of cardiovascular disease (CVD). A stroke occurs when a person’s blood vessels to the brain become blocked or burst.
There are two main types of stroke:
- i) Schemic stroke is the most common type and occurs as a result of a hemorrhagic stroke with a blocked artery
- Occurs when a blood vessel bursts and causes bleeding
There are a number of risk factors for stroke, including:
- Gender (men have a higher risk of stroke than women) Race (African Americans are 50% more likely to have a stroke
- , having a stroke) stroke (compared to white adults)
- Family anamnesis
- High blood pressure (high blood pressure)
- Pre-existing heart disease
with high LDL cholesterol levels
The symptoms of a stroke can develop either very quickly or slowly over hours or days. Typical symptoms of a stroke include:
- Numbness or weakness, particularly when only one side of the body
- unable to speak or understand, speech disorders
- , vision problems
- Dizziness and balance problems
- severe headache
The American Heart Association (AHA) recommends using the FAST test to determine symptoms:
- Q: Facial sagging
- A: arm weakness
- S: language difficulty
- T: Time to call 911
Although doctors can treat a stroke, the outcome depends on the severity of the stroke and how quickly a person receives treatment.
Using a chest X-ray to predict heart disease
According to Dr. Nicole Weinberg, a cardiologist at Providence Saint John’s Health Center in Santa Monica, California, who is not involved in the new study, many people could assume that a chest X-ray only looks at the lungs.
However, when a radiologist takes a picture of a person’s chest, he also sees other organs in the lung area, including the heart.
“If the heart is really big, if it’s kind of deformed, the radiologist can give us a feel for it,” Dr. Weinberg told Medical News Today.
“You can also see the aorta — if it is enlarged or there are calcium deposits there, it lights up. They can also see in lung tissue whether fluid is building up either at the base of the lung fields or indoors, so there may be heart failure or additional volume in the system.”
Development of a predictive model for strokes
Researchers developed a deep learning model using artificial intelligence (A) I) for the current study.
The team used a CXR-CVD system that was “trained” to search more than 147,000 chest X-ray images from nearly 41,000 participants in a cancer screening trial and to identify patterns associated with cardiovascular disease.
Once developed, the system could predict a person’s 10-year risk of having a stroke or heart attack using a single chest X-ray.
Lead author of the study, Jakob Weiss, Ph.D., a radiologist at the Cardicular Imaging Research Center at Massachusetts General Hospital and the AI in Medicine program at Brigham and Women’s Hospital in Boston, told MNT:
“Current guidelines from the American College of Cardiology and the American Heart Association for primary cardiovascular disease prevention recommend using a risk calculator to estimate the risk of future cardiovascular disease. This risk calculator is based on the ASCVD risk score, a multi-variable regression model that requires nine variables as input, such as age, sex, smoking, lipids, blood pressure, and diabetes. However, these variables are often not available, making novel and more practical screening approaches desirable.”
Dr. Weiss added that the new AI model uses a routine chest X-ray as the only input, which “allows opportunistic screening of cardiovascular risk in the event that the variables to calculate the risk score recommended by the guideline are missing.”
Next steps in AI research to detect strokes
Although compelling, the new research results are preliminary and further long-term studies are still needed.
Dr. Jennifer Wong, cardiologist and medical director for noninvasive cardiology at the Orange Coast Medical Center’s MemorialCare Heart and Vascular Institute in Fountain Valley, California, who was not involved in the study, said future longitudinal studies could look at the results of people at various risks, as revealed by a chest X-ray.
“It’s important to identify people at higher risk of cardiovascular disease early on, and more tools to identify people at risk are always useful,” Dr. Wong told MNT.
“This study did an excellent job of correlating with the current tool cardiologists are using, the ASCVD risk score, but using where they’re stratifying risk, to see what happens to that person’s risk of heart attack and stroke 10 (or) 20 years later, using just a chest X-ray as a tool could be helpful,” she added.