AI & Heart Imaging: Deep Learning for Accessibility
Breakthrough in heart health! Scientists have developed CTLESS, a pioneering deep learning method, fundamentally changing heart imaging. It removes the need for additional CT scans,dramatically reducing radiation exposure for patients undergoing myocardial perfusion imaging (MPI).This innovation promises improved access to vital heart health monitoring, especially benefiting rural areas and communities with limited resources, where such scans may be constrained.CTLESS leverages advanced algorithms to enhance image quality,delivering diagnostic accuracy comparable to customary methods while cutting costs. This work, highlighted in IEEE Transactions in Medical Imaging and potentially a game-changer, could reshape how we approach cardiac care. News Directory 3 provides insightful coverage on these technological developments. Discover what’s next in this exciting field.
Deep Learning Improves Heart Imaging, Reduces Radiation Exposure
Updated June 09, 2025
A new deep learning technique promises to revolutionize heart health monitoring by making it safer and more accessible. Researchers at Washington University in St. Louis, in collaboration with Cleveland Clinic and university of California Santa Barbara, have developed a method called CTLESS that eliminates the need for additional CT scans during myocardial perfusion imaging (MPI) by single photon emission computed tomography (SPECT).
SPECT imaging, a common tool for diagnosing coronary artery disease, traditionally requires a CT scan for attenuation compensation, wich corrects for signal weakening as it passes through body tissue. This additional scan increases radiation exposure and costs. Abhinav Jha, associate professor at WashU Medicine Mallinckrodt Institute of Radiology, led the project, which was published in IEEE Transactions in Medical Imaging.
Jha said their cost-saving technique is particularly critically important for cases where access to such scans might potentially be limited, such as in rural or or else resource-limited communities. The next stage of research is for them to validate this method while working to make this tech more available to rural community hospitals.
CTLESS uses photons from the emission scan to estimate attenuation, enhancing image quality and diagnostic interpretation. Clinical data showed CTLESS achieving results comparable to customary attenuation compensation, with robust performance across different scanner models, degrees of heart damage, and patient demographics.
“Due to cost, complexity, equipment availability, regulatory concerns and other local factors at hospitals and remote care centers, approximately 75% of all SPECT MPI scans are performed without AC, potentially compromising the diagnostic accuracy of these scans,” Jha said.
Jha also noted that CTLESS performed consistently for both men and women, despite anatomical differences affecting attenuation levels. The method’s stability, even with reduced training data, makes it a promising candidate for widespread clinical adoption after further validation. This advancement in deep learning for heart imaging could significantly improve access to accurate diagnoses, especially in underserved areas.
What’s next
The researchers plan to validate CTLESS further and work toward making the technology more accessible to rural community hospitals, potentially boosting technological health equality across the U.S. and worldwide.