Nov. 23, 2024
Metabolic syndrome, a major risk factor for cardiovascular diseases, is on the rise. More than a third of the U.S. population and a quarter of people globally are affected by the condition. What if metabolic syndrome risk in patients could be predicted taking pictures with a tablet and using an app? Mayo Clinic Cardiovascular Medicine researchers found that artificial intelligence (AI) combined with an advanced 3D body-volume scanner could help predict a patient's risk and severity of metabolic syndrome.
"Metabolic syndrome is impacting people at risk for heart disease, those with lower educational levels, socioeconomic status, and those who experience food insecurity," says Betsy Medina Inojosa, M.D., a research fellow at Mayo Clinic in Rochester, Minnesota, and first author of the study. "Recognition and accurate metabolic syndrome risk stratification are essential to guide early preventative interventions."
Using a 3D body-volume scanner, imaging and mobile technology, and Mayo Clinic-developed algorithms offer a more precise measure of risk than does BMI and waist-to-hip ratio, according to findings published in European Heart Journal — Digital Health.
Ready, reliable results
The absence of widely accepted screening strategies motivated Mayo Clinic researchers to create a solution. Plus, laboratory tests are not always readily available in certain clinical settings.
"This study emerged from the need for a reliable and easy-to-reproduce measure of metabolic syndrome risk and severity. Metabolic syndrome diagnosis is based on anthropometric measurements that in clinical practice and research have limited acceptance due to high intra-interobserver variability," says Dr. Medina Inojosa. "We previously demonstrated that a multisensor 3D body-volume scanner is a reliable and valid technique for determining body circumferences and body composition."
The scanner is a noninvasive device used in the clothing industry to assess body shape and size. A mobile BVI application also was used with biplane imaging captured with mobile technology to determine body volumes accurately.
Creating innovative methodology
Researchers trained and validated an AI model on 1,280 people. More than half were women and predominantly non-Hispanic white. "To the best of our knowledge, this is the first study to develop and validate a metabolic syndrome prediction model using biplane imaging with a body composition calculator app to determine body volumes. Other groups are now trying to replicate our work," says Francisco Lopez-Jimenez, M.D., M.S., a cardiologist, chair of Preventive Cardiology and co-director of Artificial Intelligence in Cardiology at Mayo Clinic in Minnesota and senior author of the study. "The uniqueness of our work lies in its methodology."
External validation was performed on 133 people of whom 58% were women and 74% were non-Hispanic white. The extra volunteers had front- and side-view images taken via the body composition calculator app to further test whether they had metabolic syndrome.
"This study describes a model that accurately predicts metabolic syndrome prevalence and severity, with excellent and better prediction performance compared to conventional assessment like body mass index or waist-to-hip ratio," says Dr. Lopez-Jimenez. "We implemented advanced machine learning methods to refine the predictive capabilities of this technology. Lastly, an external cohort of individuals scanned with the BVI mobile app serves as external validation, corroborating model performance across different settings and devices."
Looking ahead
Future research needs to include greater diversity to study the relationship between the BVI and cardiovascular events and the long‐term changes in body composition and cardiovascular disease prognosis.
Artificial intelligence and 3D imaging often help patients in various ways. "Digital technologies have the potential to close gaps in healthcare. The technologies used in this study provide reliable and valid techniques for determining body volumes and composition, representing a promising, portable, affordable and scalable tool with potential applicability in remote settings, low-income countries and large-scale populations while enhancing telemedicine and telecare services," says Dr. Lopez-Jimenez. "These measures may serve as opportunities to facilitate the prevention, monitoring and management of cardiometabolic conditions such as metabolic syndrome."
For more information
Medina Inojosa BJ, et al. Prediction of presence and severity of metabolic syndrome using regional body volumes measured by a multisensor white-light 3D scanner and validation using a mobile technology. European Heart Journal — Digital Health. 2024;5:582.
Refer a patient to Mayo Clinic.