For over a century, health professionals have peered into patients’ eyes to diagnose cardiovascular diseases (CVD).

The process involves looking for changes in retinal vasculature — blood vessels in the retina that reflect the impact of high blood pressure over time. It indicates the presence or absence of cardiovascular diseases.

Over time, medical scientists have developed several instruments to provide a more convenient look at the retinal vasculature.

However, these devices all have the same shortcomings — human judgment. Medical professionals still have to make the final call after peering at the patient’s eyes.

Now, an international team of scientists has taught an AI system to recognize the same symptoms for heart disease. With that, human intervention will no longer be required for diagnosis.

In their published paper in the journal Nature Biological Engineering, the researchers wrote:

“Our findings motivate the development of clinically applicable explainable end-to-end deep-learning systems for the prediction of CVD on the basis of the features of retinal vessels in retinal photographs.”

Here’s how they developed the AI system.

Developing an AI System to Diagnose Cardiovascular Disease

First, the researchers taught the AI system to recognize CVD markers using thousands of retinal vasculature scans.

They also trained the system using data that highlights the associated CVD markers for specific patients. For this part of the study, the team used over 70,000 images from people with diverse backgrounds.

After that, the researchers tested the AI system’s accuracy.

They compared the model’s diagnosis with the traditional methods to Diagnose Cardiovascular Disease. These involve using factors such as obesity, BMI, cholesterol, diabetes, and glycated hemoglobin levels.

According to the researchers, their system outperformed the health professionals based on patient tracking after testing.

Indeed, it’s not the first time that researchers are training artificial intelligence to study retinal scans.

Back in 2018, Google AI researchers taught their deep learning algorithms to predict heart disease symptoms from retinal images. What’s more, the algorithm was able to deduce other data such as age, gender, smoking status, and systolic blood pressure.

AI-based diagnostic systems can serve as a new tool to help health professionals catch CVD in patients early. That way, they’ll not only lessen the symptoms but also prevent a tragic outcome.

Read More: Owning a Dog May Improve Your Cardiovascular Health

The post New AI Model can Predict the Risk of Cardiovascular Disease appeared first on EDGY_ Labs.