Researchers have created a new AI method that can find plaque erosion in the arteries of the heart. Optical coherence tomography (OCT) images are used by the technician to look at the arterial plaque.
This finding is helpful because a heart attack or other serious heart disease can begin before the plaque breaks down. OCT can create 3D pictures of the coronary arteries, the blood vessels that carry blood to the muscles of the heart.
OCT has also been used to look for plaque erosion. But because there are so many data points and the images are hard to understand, the results are very different. AI was used by researchers to figure out how to solve this problem.
How does the AI ​​technology work?
The new method for AI consists of two main steps. First, an AI model called a neural network looks at the original image and two shapes to predict where plaque erosion will occur.
The prediction is then made even better by a post-processing algorithm that mimics the knowledge of a professional doctor to make a diagnosis.
Zhao Wang of the University of Electronic Science and Technology of China says: “We needed to create a new AI model that contains explicit shape information, which is the main feature used to find plaque erosion on OCT images. The underlying intravascular OCT imaging technology is also important because it is currently the highest-resolution imaging modality that can be used to diagnose plaque erosion in living patients.
Researchers tested the new method and found that the automated method could predict 80% of plaque erosion cases with a positive predictive value of 73.3%. Also, the AI’s diagnoses were the same as those of doctors with years of experience.
Zhao said, “Our new method could help improve the diagnosis of plaque erosion in the clinic and could be used to create new treatments for people with heart disease.”