The emergence of artificial intelligence (AI) is transforming cardiovascular medicine. Initially, AI applications concentrated on analyzing single data types, such as electrocardiograms and imaging studies. However, advancements in multimodal AI have now enabled the integration of diverse data sources, facilitating a comprehensive understanding of patient health and predictive accuracy of disease outcomes. In this review, we discuss current achievements in multimodal AI within cardiovascular medicine, including various combinations of different modalities, computer algorithms of data integration and fusion, and their integration into clinical workflow. As the field continues to evolve, we further propose current challenges and prospects for their future implementation.
Objectives To evaluate changes in health outcomes between years 2 and 3 after discharge following COVID-19 and to identify risk factors for poor health 3-year post-discharge. Design This is a multicentre observational cohort study. Setting This study was conducted in two centres from Wuhan, China. Participants Eligibility screening has been performed in 3988 discharged laboratory-confirmed adult COVID-19 patients. Exclusion criteria were refusal to participate, inability to contact and death before follow-up. The WHO COVID-19 guidelines on defining disease severity were adopted. Results 1594 patients participated in the 1-year, 2-year and 3-year follow-ups, including 796 (49.9%) male patients, and 422 (26.5%) patients were classified in the severe disease group. 3 years after discharge, 182 (11.4%) patients still complained of at least one symptom. The most common symptoms were fatigue, myalgia, chest tightness, cough, anxiety, shortness of breath and expectoration. Fatigue or myalgia, the most common symptom cluster, frequently coexisted with chest symptoms and anxiety. Symptom persistence between years 2 and 3 was reported in 70 patients (4.4%) for which intensive care unit (ICU) admission was a risk factor (p=0.038). Of the 1586 patients who completed the chronic obstructive pulmonary disease assessment test (CAT), 97 (6.1%) scored ≥10, with older age being associated with CAT ≥10 (p=0.007). Conclusions Between years 2 and 3 after SARS-CoV-2 infection, most patients returned to an asymptomatic state, and only a few were still symptomatic. ICU admission was a risk factor for symptom persistence.
15 December 2025 in JACC: Cardiovascular Interventions
Background/Objectives: Transcatheter aortic valve implantation (TAVI) in patients with bicuspid aortic valve (BAV) remains associated with higher rates of residual paravalvular leak (PVL), which confers a two-fold increase in mortality. Despite procedural optimization including balloon post-dilatation, a subset of patients exhibit residual ≥moderate PVL. Pre-procedural identification of these patients could guide procedural planning. Methods: We retrospectively analyzed 402 BAV patients who underwent TAVI with self-expanding valves and balloon post-dilatation between January 2016 and June 2024. A multi-modal deep learning model (Model B) was developed, integrating a 3D ResNet encoder for computed tomography (CT) imaging features with a multilayer perceptron (MLP) for clinical variables, fused via a cross-attention mechanism. Its performance was compared against a conventional model (Model A) combining clinical variables with manually derived CT measurements. Both models were evaluated on identical test folds using 5-fold stratified cross-validation. Results: Of 402 patients, 36 (9.0%) had residual ≥moderate PVL, associated with significantly larger aortic root dimensions at all anatomical levels and greater aortic valve calcification volume (median 887.6 vs. 559.2 mm3; p = 0.004). Model A achieved a mean AUC of 0.694 (95% CI: 0.596–0.792). Model B achieved a mean AUC of 0.822 (95% CI: 0.680–0.964), with a specificity of 0.971, accuracy of 0.881, and PPV of 0.860, while sensitivity was 0.429, reflecting the limited number of outcome events in this cohort. Conclusions: A multi-modal deep learning model integrating expert-segmented CT imaging with clinical variables demonstrated significantly improved discrimination over the conventional approach in this internal cohort for predicting residual PVL in BAV-TAVI, supporting the integration of segmentation-guided deep learning into pre-procedural TAVI planning. However, given the modest number of outcome events, external validation is required to confirm the generalizability of these findings.
Background/Objectives: Post-procedural permanent pacemaker implantation (PPMI) and cardiac damage are individually associated with adverse outcomes following transcatheter aortic valve replacement (TAVR). However, their joint impact has not been systematically evaluated. Methods: Individuals who underwent TAVR procedures between 2013 and 2024 were retrospectively enrolled and classified into four groups based on their PPMI and cardiac damage status. Univariate and multivariable Cox regression were used to analyze the association of these factors with long-term outcomes, while receiver operating characteristic (ROC) curves, decision curve analysis (DCA), net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were used to further assess their predictive performance. Results: A total of 1274 patients met the inclusion criteria. Compared to non-PPMI patients with early cardiac damage stage, those with PPMI and advanced stage had higher all-cause mortality (HR: 2.98; 95% CI: 1.60–5.56; p value < 0.001) and more cardiac deaths (HR: 3.85; 95% CI: 1.36–11.10; p = 0.012). Multivariable analysis confirmed PPMI and cardiac damage stage as independent prognostic factors. Notably, the predictive model incorporating both variables achieved the best performance (AUC = 0.703) with significant NRI and favorable DCA results, suggesting incremental value for identifying high-risk patients and acceptable clinical utility. Conclusions: The current study is among the first to explore whether post-procedural cardiac damage may identify a subgroup in whom PPMI carries a particularly adverse long-term prognosis, highlighting the need for integrated risk assessment for post-TAVR lifelong management.
24 August 2026 in JACC: Cardiovascular Interventions