A team led by researchers at Queen Mary University of London has developed new AI tools capable of automatically measuring the amount of fat around the heart based on MRI scans. The scientists were thus able to show that a...Fatty heart disease with a significantly higher risklinked to the development of diabetes. This method was carried out regardless of the age, gender and body mass index of the person examined.
Be able to make quick diagnoses using AI tools
The distribution of fat in the body can affect a person's risk of various diseases. However, the commonly used measure of body mass index reflects the accumulation of fat under the skin, rather than around the internal organs. In particular, there is evidence that fatty heart disease can be a predictor of heart disease. Doctors link these to a number of diseases, including atrial fibrillation, diabetes and coronary heart disease. Unfortunately, manually measuring the amount of fat around the heart is a time-consuming challenge. For this reason, no one has been able to examine this thoroughly in studies of large groups of people. The AI tools apply standardized MRI scans of the heart. In this way, the researchers automatically and quickly obtained a measurement of the fatty tissue around the heart in less than three seconds. This way you can find out more about the connections between fatty heart disease and the risk of disease. This could also become part of a patient's standard care in the hospital in the future.
The research team tested the algorithm's ability to interpret images from cardiac MRI scans of more than 45,000 people, including a database of health information from over half a million participants. The scientists found that the AI tools could accurately determine the amount of fat around the heart in these images and also calculate a patient's risk of diabetes. The whole thing also includes a built-in method for calculating the uncertainty of its own results. This enables his impressive ability to mark his own values. The inthis studyThe novel AI tools presented would therefore be of great benefit for future research. If proven, the method can be used in clinical practice to improve patient care, particularly in cardiovascular imaging.
