Radiomics Model Predicts Heart Attacks From CT Images

Large, lipid-rich plaques are especially prone to rupturing which ultimately leads to most heart attacks. Identifying which plaques may rupture in the future has so far been difficult. To address this issue, researchers are now using an approach called radiomics that can generate quantitative data from CT scans in order to detect underlying characteristics that are not normally visible in the images. This data can be used to predict future cardiac events such as heart attacks.

Researchers at the Medical School of Nanjing University (Nanjing, China) have created a radiomics model that utilizes data from coronary CT angiography scans to evaluate plaque vulnerability. The model was developed using data from 299 patients and then tested in 708 patients suspected of having coronary artery disease. The model enabled the researchers to identify plaques that carried a higher risk of major adverse cardiac events such as heart attacks. Furthermore, a high radiomic signature was found to be independently associated with these events over a median three-year follow-up.

According to the researchers, the radiomic signature can be easily added to clinical practice. It could assess plaques that are possibly vulnerable and aid in distinguishing high-risk patients in the clinic. The researchers next intend to develop a radiomics model from different scanner types and vendors and also plan to conduct a larger, multicenter study of 10,000 patients.

"Radiomics provided a more accurate approach to detect vulnerable plaques compared to conventional coronary CT angiography anatomical parameters," said study co-lead author Long Jiang Zhang, M.D., Ph.D., from the Department of Radiology at Jinling Hospital, Medical School of Nanjing University in Nanjing. "If the radiomics analysis is embedded into the routine CT angiography workstation, it can automatically identify vulnerable plaques for clinician review. Thus, radiomics may significantly improve the accuracy and precision of high-risk plaque detection in routine clinical practice."

Related Links:
Nanjing University