Myocardial focal scar (MFS) is a pathological state widely found in ischemic and non-ischemic cardiomyopathy. It is a muti-complication pathological tissue with poor prognosis and high mortality, especially for left ventricular old myocardial infarction (OMI) [1]. With the left ventricular remodeling of OMI,various complications owing to MFS, such as ventricular aneurysm, left ventricular enlargement, cardiac function decreasing, and malignant arrhythmia, affect patients' quality of life and increase the risk of major adverse cardiovascular events (MACE), including heart failure and sudden cardiac death [2].

MFS is usually detected by late gadolinium enhancement (LGE) on cardiac magnetic resonance imaging (CMR), which is technically difficult, economically inefficient and promoted limited [3,4]. Coronary computed tomography angiography (CCTA) is now a first-line non-invasive investigation for assessing possible coronary artery disease (CAD) [5]. MFS generally appears as lower density in CCTA images compared with normal myocardium. However, given the inherent limitations of CT images concerning tissue resolution and beam hardening (BH) which mimic MFS as low density and significantly impeding the diagnostic accurate, CCTA is not currently used to detect MFS. Dual-energy CCTA (DECCTA) is a special form of CCTA that employs two X-ray tubes or one X-ray tube with rapid kV-switching to generate virtual monochromatic images (VMIs). This technique has emerged as a valuable tool, offering the potential to enhance contrast between scar and normal myocardium while simultaneously reducing the impact of beam hardening (BH) [[6], [7], [8], [9], [10], [11]].

Radiomics has been proven as an effective methods to improve the diagnostic efficiency of cardiac diseases, such as cardiomyopathy and myocardial infarction [12,13]. It is postulated that radiomics has the potential to facilitate a comprehensive investigation of the intrinsic data within diverse DECCTA VMIs. However, none of the previous studies about improving diagnostic efficiency of cardiac diseases through radiomics was based on VMIs of DECCTA.

We suppose radiomics may further improve the differential diagnostic efficacy between BH and MFS by discerning concealed features in VMIs of DECCTA, so we seek to devise and validate a DECCTA-based radiomics model for the accurate identification of MFS and BH.