The integration of computational technologies into cardiology has significantly advanced the diagnosis and management of cardiovascular diseases. Computational cardiology, particularly through cardiovascular imaging and informatics, enables precise diagnosis of myocardial diseases by utilizing techniques such as echocardiography, cardiac magnetic resonance imaging, and computed tomography. Early-stage disease classification, especially in asymptomatic patients, benefits from these advancements, potentially altering disease progression and improving patient outcomes. Automatic segmentation of myocardial tissue using Deep Learning (DL) algorithms improves efficiency and consistency in analyzing large patient populations. Radiomic analysis can reveal subtle disease characteristics from medical images and can enhance disease detection, enable patient stratification, and facilitate monitoring of disease progression and treatment response. Radiomic biomarkers have already demonstrated high diagnostic accuracy in distinguishing myocardial pathologies and promise treatment individualization in cardiology, earlier disease detection, and disease monitoring. In this context, this narrative review explores the current state of the art in DL applications in medical imaging (CT, CMR, echocardiography and SPECT), focusing on automatic segmentation, radiomic feature phenotyping, and prediction of myocardial diseases, while also discussing challenges in integration of DL models in the clinical practice.

Background: Cardiomyopathies are a significant cause of heart failure, arrhythmia, and cardiac morbidity in the general population. Cardiovascular magnetic resonance (CMR) is a valuable tool for the diagnostic work-up of patients with acute cardiac events. Objectives: This study evaluated the diagnostic value of CMR and the yield of cardiomyopathies in hospitalized cardiac patients with acute presentation. Methods: A retrospective analysis was conducted with 535 consecutive hospitalized patients who underwent CMR at Hippokration Hospital, Athens, Greece, to identify a subset of scans performed on an urgent basis of hospitalized patients. Demographic data, causes of admission, CMR findings, and plasma cardiac biomarkers (hs-Troponin I, NT-proBNP, and CRP) were systematically recorded. Results: Out of the initial 535 CMR scans evaluated, a further analysis was conducted with 104 patients who were in hospital and underwent CMR on an urgent basis. From the total population of hospitalized patients, 33% had CMR findings indicative of underlying cardiomyopathy, with dilated cardiomyopathy being the most common subtype (36%), followed by arrhythmogenic cardiomyopathy (27%), hypertrophic cardiomyopathy (15%), or other subtypes (e.g., cardiac amyloidosis, sarcoidosis, endomyocardial fibrosis, EGPA, or unclassified). CMR led to the reclassification of the initial diagnosis into that of underlying cardiomyopathy in 32% of cases. The highest reclassification rate was observed within the subgroup with heart failure (71%), followed by that of acute myocardial infarction/ischemic heart disease (24%) and myocarditis (22%). Conclusions: CMR imaging effectively contributed to the differential diagnosis of hospitalized patients with acute cardiac events that remained without a definitive diagnosis after their initial work-up and uncovered underlying cardiomyopathy in almost one-third of this cohort.

Background Cardiomyopathies are a significant cause of heart failure, arrhythmia and cardiac morbidity in the general population/ Cardiovascular magnetic resonance (CMR) is valuable tool for the diagnostic work-up of patients with acute cardiac events. Objectives: This study evaluated the diagnostic value of CMR and the yield of cardiomyopathies in hospitalized cardiac patients with acute presentation. Methods A retrospective analysis was conducted in 535 consecutive hospitalized patients who underwent CMR at Hippokration Hospital, Athens, Greece, to identify a subset of scans performed at an urgent basis of hospitalized patients. Demographic data, cause of admission, CMR findings, and plasma cardiac biomarkers (hs-Troponin I, NT-proBNP, CRP) were systematically recorded. Results Out of the initial 535 CMR scans evaluated, further analysis was conducted on 104 patients who were in-hospital and underwent CMR on an urgent basis. From the total population of hospitalized patients, 33% had CMR findings indicative of an underlying cardiomyopathy, with dilated cardiomyopathy being the most common subtype (36%) followed by arrhythmogenic cardiomyopathy (27%), hypertrophic cardiomyopathy (15%) or other subtypes (e.g., cardiac amyloidosis, sarcoidosis, endomyocardial fibrosis, EGPA, or unclassified). CMR led to reclassification of the initial diagnosis into that of an underlying cardiomyopathy in 32% of cases. The highest reclassification rate was observed within the subgroup with heart failure (71%), followed by that of acute myocardial infarction/ischemic heart disease (24%) and myocarditis (22%). Conclusions CMR imaging effectively contributed to the differential diagnosis of hospitalized patients with acute cardiac events that remained without a definitive diagnosis after their initial work-up and uncovered underlying cardiomyopathy in almost one-third of this cohort.

Transthyretin cardiac amyloidosis (ATTR) is prevalent among patients undergoing transcatheter aortic valve implantation (TAVI) for severe aortic stenosis (AS). We aimed to develop and validate a risk score for predicting ATTR in this context. In the prospective multicentre GRECA-TAVI registry, 500 patients with severe AS scheduled for TAVI were screened for ATTR using technetium-99m-labeled DPD or PYP bone scintigraphy in 12 tertiary hospitals, from January 1st to December 31st, 2024. Clinical, echocardiographic, and electrocardiographic variables were recorded. A logistic regression-derived integer-based risk score was developed in a training cohort (n=350) and validated in a validation cohort (n=150). Model performance was evaluated using discrimination, calibration, and decision curve analyses. ATTR was diagnosed in 38 patients (7.6%). Five variables (age, sex, MWT, LVEF, and intracardiac device presence) formed the basis of an 8-point risk score. The score demonstrated good discrimination (AUC=0.75 training, 0.76 validation) and calibration (slope=1.15). A score ≥4 identified patients with a ∼16% ATTR prevalence compared to 2.4% in those with a score <4. This threshold captured most of the ATTR cases (73%) while recommending screening for only 36% of the cohort. Decision curve analysis showed net clinical benefit across clinically relevant thresholds (5-20%) over and above a guideline-recommended approach (based on MWT≥12mm) or a screen-all strategy. ATTR is common in patients with severe AS undergoing TAVI. A simple bedside score using routinely available variables effectively identifies those patients at risk for ATTR. Selective screening using this tool could improve diagnostic yield and resource allocation.

Diagnosis of cardiomyopathies often depends on overt phenotypic manifestations, delaying patient management. This underscores the need for population-level opportunistic screening tools using clinically indicated CT scans to detect subclinical myocardial disease. This study developed an Ensemble Machine Learning (ML) model to automatically segment the left ventricular myocardium from CT data and estimate the probability of underlying myocardial disease using radiomic feature analysis. A total of 60 CT scans (~12,000 images) were used to train ML models for left ventricular myocardium segmentation, including scans from both healthy individuals and patients with myocardial disease. A novel Ensemble model was developed and externally validated on 10 independent CT scans. Subsequently, 100 unseen CT scans were segmented manually and automatically for radiomic feature analysis. After removing highly correlated features through intra-class variation and correlation filtering, the refined dataset was used for model training and testing. Key predictive features were identified, and model performance was evaluated. The four best-performing models (Unet++, ED w/ASC, FPN, and TresUNET) were combined to form an Ensemble model, achieving a final DICE score of 0.882 after hyperparameter optimization. External validation yielded a DICE score of 0.907. Radiomic feature analysis identified 15 key predictors of myocardial disease in both manual and automatic segmentation datasets. The model demonstrated strong performance in detecting underlying myocardial disease, with AUCs of 0.85 and 0.8, respectively. This study presents a fully automated CT-based framework for LV myocardial segmentation and radiomic phenotyping that accurately estimates the probability of underlying myocardial disease. The model demonstrates strong generalizability across different CT protocols and highlights the potential role of AI-driven CT analysis for early, non-invasive cardiomyopathy screening at a population level.

Background: Diagnoses of transthyretin amyloid cardiomyopathy (ATTR-CM) have increased in recent years, with a growing proportion of patients being identified at earlier disease stages. However, temporal trends in ATTR-CM diagnosis and survival during the tafamidis era in the Greek population remain largely unknown. Methods: This retrospective study included 75 patients with ATTR-CM treated with tafamidis. Baseline demographic, clinical, laboratory, and echocardiographic characteristics were assessed, and survival outcomes were subsequently evaluated. Results: Among 75 tafamidis-treated patients, 14 deaths (19%) occurred during a median follow-up of 20 months [IQR 10.0–30.5] with an estimated 30-month survival of 75%. Kaplan–Meier analysis demonstrated significantly poorer survival among patients aged ≥80 years (log-rank p = 0.013), those with NYHA class III–IV (log-rank p = 0.018), and across progressively more advanced NAC stages I–III (global log-rank p < 0.001). In contrast, survival did not differ significantly according to LVEF <50% versus ≥50% (log-rank p = 0.863) or maximal wall thickness ≥15 mm versus <15 mm (log-rank p = 0.801). After Holm correction across five comparisons, only NAC stage remained statistically significant (adjusted p < 0.001). Conclusions: In this real-world Greek cohort of patients with ATTR-CM treated with tafamidis, the estimated 30-month all-cause mortality was 25% and more advanced NAC stage was associated with poorer survival.