Connect with us

Healthcare

Researchers Unravel Mysteries of the Heart and Predict Heart Disease With AI

mm

Updated

 on

Recently, two studies involving the heart have made use of AI algorithms to gain a better understanding of how the heart functions and malfunctions. A team of researchers has used artificial intelligence algorithms to gain insight into how the shape of the heart’s muscles impacts its performance. With the aid of AI algorithms, the research team gained insight into how the heart’s ventricles facilitate blood flow and found six different DNA sections that play important roles in the development of heart muscles. Meanwhile, another study examined how selfies could potentially be used to diagnose heart disease.

As far back as the 15th and 16th centuries, scientists have studies the heart and wondered how its structure was related to its function. Leonardo da Vinci wondered how heart muscles moved blood throughout the body over 500 years ago. Thanks to a team of researchers from institutes like the MRC London Institute of Medical Sciences, the Cold Spring Harbor Laboratory, EMBL’s European Bioinformatics Institute (EMBL-EBI), Heidelberg University, and the Politecnico di Milano, we are much closer to understanding the role that the heart’s trabeculae play in the development and function of the heart.

Trabeculae are muscle fibers that form a complex network of geometric patterns on the inner surface of the heart. The trabeculae are believed to provide oxygen to the heart as it develops, as the heart itself is the first organ to develop and cannot get oxygen from the lungs. However, it has long been a mystery what role the trabeculae played in adults. Da Vinci had speculated that the heart’s trabeculae served to warm the blood as it moved through the heart, but thanks to AI-based research techniques we now have an idea of their true purpose.

The research team employed AI algorithms to analyze approximately 25000 MRI (magnetic resonance imaging) scans taken of the heart. These scans were fed to the AI model alongside genetic data and heart morphology data. The researchers analyzed the model’s results and found that the trabeculae seemed to play a role in facilitating blood flow through the ventricles of the heart. The geometric patterns on the inside of the heart are now believed to help blood flow more efficiently as the heart beats.

In addition to the finding that trabeculae could assist in the efficient flow of blood, the researchers also discovered six sections of DNA that seem to impact the development of trabeculae. Two of the six DNA sections also play roles in the development of branching nerve pathways within the brain. It’s possible that the similar mechanisms that give rise to trabeculae also give rise to nerve cells.

The shape of the trabeculae may even have a relationship with heart disease. Genetic data from over 50000 patients was analyzed by the research team, and it was found that the different trabeculae patterns had some correlation with one’s risk of developing heart disease.

In a different study, as reported by Futurism, researchers from the National Center for Cardiovascular Diseases in Beijing, China experimented with an AI capable of predicting an individual’s chance of heart disease based on physical attributes that can be captured in a simple picture of a person. There are physical attributes that have correlations with heart disease, such as yellow deposits near a person’s eyelids, white rings in the cornea’s outer edges and graying or thinning hair. When trained on images of patients from Chinese hospitals, the algorithm reportedly outperformed existing heart disease risk assessment methods, being able to detect roughly 80% of heart disease cases. It also captured approximately 60% of all negative heart disease instances.

There is a high false-positive rate associated with the algorithm, which the team states that they will need to address, noting that false positives could cause patients unnecessary anxiety and overload the medical system with unneeded tests. However, if the algorithm’s false positive rate can be addressed and overall accuracy improved, it could be a useful tool for areas of the world that struggle with adequately funding cardiovascular disease screening programs.

Spread the love