Outcomes According to Coronary Disease Complexity and Optimal Thresholds to Guide Revascularization Approach: FAME 3 Trial.

BACKGROUND: Coronary disease complexity is commonly used to guide revascularization strategy in patients with multivessel disease (MVD). OBJECTIVES: The aim of this study was to assess the interactive effects of coronary complexity on percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) outcomes and identify the optimal threshold at which PCI can be considered a reasonable option. METHODS: A total of 1,444 of 1,500 patients with MVD from the FAME (Fractional Flow Reserve versus Angiography for Multi-vessel Evaluation) 3 randomized trial were included in the analysis (710 CABG vs 734 PCI). SYNTAX (Synergy Between PCI With Taxus and Cardiac Surgery) scores were transformed into restricted cubic splines, and logistic regression models were fitted, with multiplicative interaction terms for revascularization strategy. Optimal thresholds at which PCI is a reasonable alternative to CABG were determined on the basis of Cox regression model performance. RESULTS: The mean SYNTAX score (SS) was 25.9 ± 7.1. SS was associated with 1-year major adverse cardiac and cerebrovascular events among PCI patients and 3-year death, myocardial infarction, and stroke among CABG patients. Significant interactions were present between revascularization strategy and SS for 1- and 3-year composite endpoints (P for interaction <0.05 for all). In Cox regression models, outcomes were comparable between CABG and PCI for the 3-year primary endpoint for SS ≤24 (P = 0.332), with 44% of patients below this threshold and 32% below the conventional SS threshold of ≤22. CONCLUSIONS: In patients with MVD without left main disease, PCI and CABG outcomes remain comparable up to SS values in the mid- rather than low 20s, which allows the identification of a greater proportion of patients in whom PCI may be a reasonable alternative to CABG.

A Responsible Framework for Applying Artificial Intelligence on Medical Images and Signals at the Point of Care: The PACS-AI Platform.

The potential of artificial intelligence (AI) in medicine lies in its ability to enhance clinicians' capacity to analyse medical images, thereby improving diagnostic precision and accuracy and thus enhancing current tests. However, the integration of AI within health care is fraught with difficulties. Heterogeneity among health care system applications, reliance on proprietary closed-source software, and rising cybersecurity threats pose significant challenges. Moreover, before their deployment in clinical settings, AI models must demonstrate their effectiveness across a wide range of scenarios and must be validated by prospective studies, but doing so requires testing in an environment mirroring the clinical workflow, which is difficult to achieve without dedicated software. Finally, the use of AI techniques in health care raises significant legal and ethical issues, such as the protection of patient privacy, the prevention of bias, and the monitoring of the device's safety and effectiveness for regulatory compliance. This review describes challenges to AI integration in health care and provides guidelines on how to move forward. We describe an open-source solution that we developed that integrates AI models into the Picture Archives Communication System (PACS), called PACS-AI. This approach aims to increase the evaluation of AI models by facilitating their integration and validation with existing medical imaging databases. PACS-AI may overcome many current barriers to AI deployment and offer a pathway toward responsible, fair, and effective deployment of AI models in health care. In addition, we propose a list of criteria and guidelines that AI researchers should adopt when publishing a medical AI model to enhance standardisation and reproducibility.

Revolutionising Acute Cardiac Care With Artificial Intelligence: Opportunities and Challenges.

This article reviews the application of artificial intelligence (AI) in acute cardiac care, highlighting its potential to transform patient outcomes in the face of the global burden of cardiovascular diseases. It explores how AI algorithms can rapidly and accurately process data for the prediction and diagnosis of acute cardiac conditions. The review examines AI's impact on patient health across various diagnostic tools such as echocardiography, electrocardiography, coronary angiography, cardiac computed tomography, and magnetic resonance imaging, discusses the regulatory landscape for AI in health care, and categorises AI algorithms by their risk levels. Furthermore, it addresses the challenges of data quality, generalisability, bias, transparency, and regulatory considerations, underscoring the necessity for inclusive data and robust validation processes. The review concludes with future perspectives on integrating AI into clinical workflows and the ongoing need for research, regulation, and innovation to harness AI's full potential in improving acute cardiac care.

Maximising Large Language Model Utility in Cardiovascular Care: A Practical Guide.

Large language models (LLMs) have emerged as powerful tools in artificial intelligence, demonstrating remarkable capabilities in natural language processing and generation. In this article, we explore the potential applications of LLMs in enhancing cardiovascular care and research. We discuss how LLMs can be used to simplify complex medical information, improve patient-physician communication, and automate tasks such as summarising medical articles and extracting key information. In addition, we highlight the role of LLMs in categorising and analysing unstructured data, such as medical notes and test results, which could revolutionise data handling and interpretation in cardiovascular research. However, we also emphasise the limitations and challenges associated with LLMs, including potential biases, reasoning opacity, and the need for rigourous validation in medical contexts. This review provides a practical guide for cardiovascular professionals to understand and harness the power of LLMs while navigating their limitations. We conclude by discussing the future directions and implications of LLMs in transforming cardiovascular care and research.

Novel Artificial Intelligence Applications in Cardiology: Current Landscape, Limitations, and the Road to Real-World Applications.

Cardiovascular diseases are the leading cause of death globally and contribute significantly to the cost of healthcare. Artificial intelligence (AI) is poised to reshape cardiology. Using supervised and unsupervised learning, the two main branches of AI, several applications have been developed in recent years to improve risk prediction, allow large-scale analysis of medical data, and phenotype patients for personalized medicine. In this review, we examine the key advances in AI in cardiology and its limitations regarding bias in the data, standardization in reporting, data access, and model trust and accountability in cases of error. Finally, we discuss implementation methods to unleash AI's potential in making healthcare more accurate and efficient. Several steps need to be followed and challenges overcome in order to successfully integrate AI in clinical practice and ensure its longevity.

Association Between the Access Site for Coronary Angiography and Catheter-induced Coronary Artery Dissection.

Background: Catheter-induced coronary artery dissection (CICAD) is a rare complication of coronary angiography. The association between access site and CICAD remains unclear; however, transradial access (TRA) may be associated with a higher incidence of CICAD due to access vessel tortuosity and the mechanical disadvantage of catheters designed for the transfemoral access (TFA) approach. Methods: In this retrospective study, the reports of consecutive left heart catheterizations between April 2007, and December 2021 were reviewed for CICAD. Patients were excluded if the procedural report did not report an arterial access site. Identified CICAD cases were reviewed in detail. Results: There were 142/89,876 (0.16%) identified cases of CICAD. The access site was not associated with an increased risk of CICAD (0.18% with TRA vs 0.15% with TFA; relative risk [RR], 1.18; 95% CI, 0.84-1.65; P = .34) over the entire study period. With respect to TRA-related CICAD, male sex was associated with a decreased risk of dissection (RR, 0.64; 95% CI, 0.41-0.99; P = .04), but ST-elevation myocardial infarction at presentation was associated with an increased risk (RR, 3.01; 95% CI, 1.86-4.85; P < .01). In the TFA-predominant era, TRA was associated with an increased risk of CICAD (0.48% TRA vs 0.11% TFA; RR, 3.42; 95% CI, 2.05-5.69; P < .01)-an association that was not present in the TRA-predominant era. In-hospital mortality in patients with CICAD was 8.5%. Conclusions: CICAD is a rare complication of coronary angiography. Over a 15-year period, we did not demonstrate an association between access site and an increased risk of CICAD. There is substantial mortality associated with CICAD.

Aortic Annulus S-Curve: Implications for Transcatheter Aortic Valve Replacement and Related Procedures, Part 1.

Most transcatheter aortic valve replacement-related procedures (eg, transcatheter aortic valve replacement implantation depth, commissural alignment, coronary access, bioprosthetic or native aortic scallop intentional laceration to prevent iatrogenic coronary artery obstruction, paravalvular leak closure) require an optimal fluoroscopic viewing angle located somewhere along the aortic annulus S-curve. Chamber views, coronary cusp and coronary anatomy, can be understood along the aortic annulus S-curve. A better understanding of the optimal fluoroscopic viewing angles along the S-curve may translate into increased operator confidence and improved safety and efficacy while reducing procedural time, radiation dose, contrast volume, and complication rates.

Should they stay, or should they go: Do we need to remove the old cardiac implantable electronic device if a new system is required on the contralateral side?

Background: Ipsilateral approach in patients requiring cardiac implantable electronic device (CIED) revision or upgrade may not be feasible, primarily due to vascular occlusion. If a new CIED is implanted on the contralateral side, a common practice is to explant the old CIED to avoid device interaction. Objective: The purpose of this study was to assess a conservative approach of abandoning the old CIED after implanting a new contralateral device. Methods: We used an artificial intelligence algorithm to analyze postimplant chest radiographs to identify those with multiple CIEDs. Outcomes of interest included device interaction, abandoned CIED elective replacement indicator (ERI) behavior, subsequent programming changes, and explant of abandoned CIED. Theoretical risk of infection with removal of abandoned CIED was estimated using a validated scoring system. Results: Among 12,045 patients, we identified 40 patients with multiple CIEDs. Occluded veins were the most common indication for contralateral implantation (n = 27 [67.5%]). Fifteen abandoned CIEDs reached ERI, with 4 reverting to VVI 65. One patient underwent explant due to device interaction, and 2 required device reprogramming. Of 32 patients with an implantable cardioverter-defibrillator, 8 (25%) had treated ventricular arrhythmia. There were no failed or inappropriate therapies due to interaction. Eighteen patients (45%) had hypothetical >1% annual risk of hospitalization for device infection if the abandoned CIED had been explanted. Conclusion: In patients requiring new CIED implant on the contralateral side, abandoning the old device is feasible. This approach may reduce the risk of infection and concerns regarding abandoned leads and magnetic resonance imaging scans. Knowledge of ERI behavior is essential to avoid device interactions.

The Evolving Role of Artificial Intelligence in Cardiac Image Analysis.

Research in artificial intelligence (AI) has progressed over the past decade. The field of cardiac imaging has seen significant developments using newly developed deep learning methods for automated image analysis and AI tools for disease detection and prognostication. This review is aimed at those without special background in AI. We review AI concepts and survey the growing contemporary applications of AI for image analysis in echocardiography, nuclear cardiology, cardiac computed tomography, cardiac magnetic resonance, and invasive angiography.

Implications of Myocardial Infarction on Management and Outcome in Cardiogenic Shock.

Background The randomized DOREMI (Dobutamine Compared to Milrinone) clinical trial evaluated the efficacy and safety of milrinone and dobutamine in patients with cardiogenic shock. Whether the results remain consistent when stratified by acute myocardial infarction remains unknown. In this substudy, we sought to evaluate differences in clinical management and outcomes of acute myocardial infarction complicated by cardiogenic shock (AMICS) versus non-AMICS. Methods and Results Patients in cardiogenic shock (n=192) were randomized 1:1 to dobutamine or milrinone. The primary composite end point in this subgroup analysis was all-cause in-hospital mortality, cardiac arrest, non-fatal myocardial infarction, cerebrovascular accident, the need for mechanical circulatory support, or initiation of renal replacement therapy (RRT) at 30-days. Outcomes were evaluated in patients with (n=65) and without (n=127) AMICS. The primary composite end point was significantly higher in AMICS versus non-AMICS (hazard ratio [HR], 2.21; 95% CI, 1.47-3.30; P=0.0001). The primary end point was driven by increased rates of all-cause mortality, mechanical circulatory support, and RRT. No differences in other secondary outcomes including cardiac arrest or cerebrovascular accident were observed. AMICS remained associated with the primary composite outcome, 30-day mortality, and RRT after adjustment for age, sex, procedural contrast use, multivessel disease, and inotrope type. Conclusions AMI was associated with increased rates of adverse clinical outcomes in cardiogenic shock along with increased rates of mortality and initiation of mechanical circulatory support and RRT. Contrast administration during revascularization likely contributes to increased rates of RRT. Heterogeneity of outcomes in AMICS versus non-AMICS highlights the need to study interventions in specific subgroups of cardiogenic shock. Registration URL: https://www.clinicaltrials.gov; Unique identifier: NCT03207165.

Percutaneous Closure of a Giant Aortic Pseudoaneurysm Using Multimodality Imaging Guidance.

Ascending aortic pseudoaneurysm is a rare, life-threatening complication of cardiac surgery. Surgical management is recommended, however, transcatheter techniques offer a less invasive alternative. We describe successful percutaneous closure, guided by using multimodality imaging, in a patient with high surgical risk.

Double S-Curve Versus Cusp-Overlap Technique: Defining the Optimal Fluoroscopic Projection for TAVR With a Self-Expanding Device.

OBJECTIVES: The purpose of this study was to assess the concordance between transcatheter aortic valve implantation angles generated by the "double S-curve" and "cusp-overlap" techniques. BACKGROUND: The "double S-curve" and "cusp-overlap" methods aim to define optimal fluoroscopic projections for transcatheter aortic valve replacement (TAVR) with a self-expandable device. METHODS: The study included 100 consecutive patients undergoing TAVR with self-expanding device planned by multidetector computed tomography. TAVR was performed using the double S-curve model, as a view in which both the aortic valve annulus and delivery catheter planes are displayed perpendicularly on fluoroscopy. Optimal projection according to the cusp-overlap technique was retrospectively generated by overlapping the right and left cups on the multidetector computed tomography annular plane. The angular difference between methods was assessed in spherical 3 dimensions and on the left and right anterior oblique (RAO) and cranial and caudal (CAU) axes. RESULTS: The double S-curve and cusp-overlap methods provided views located in the same quadrant, mostly the RAO and CAU, in 92% of patients with a median 3-dimensional angular difference of 10.0° (interquartile range: 5.5° to 17.9°). The 3-dimensional deviation between the average angulation obtained by each method was not statistically significant (1.49°; p = 0.349). No significant differences in average coordinates were noted between the double S-curve and cusp-overlap methods (RAO: 14.7 ± 15.2 vs. 12.9 ± 12.5; p = 0.36; and CAU: 27.0 ± 9.4 vs. 26.9 ± 10.4; p = 0.90). TAVR using the double S-curve was associated with 98% device success, low complication rate, and absence of moderate-to-severe paravalvular leak. CONCLUSIONS: The double S-curve and cusp-overlap methods provide comparable TAVR projections, mostly RAO and CAU. TAVR using the double S-curve model is associated with a high rate of device success and low rate of procedural complications.

Optimal Fluoroscopic Projections of Coronary Ostia and Bifurcations Defined by Computed Tomographic Coronary Angiography.

OBJECTIVES: The aim of this study was to define the optimal fluoroscopic viewing angles of both coronary ostia and important coronary bifurcations by using 3-dimensional multislice computed tomographic data. BACKGROUND: Optimal fluoroscopic projections are crucial for coronary imaging and interventions. Historically, coronary fluoroscopic viewing angles were derived empirically from experienced operators. METHODS: In this analysis, 100 consecutive patients who underwent computed tomographic coronary angiography (CTCA) for suspected coronary artery disease were studied. A CTCA-based method is described to define optimal viewing angles of both coronary ostia and important coronary bifurcations to guide percutaneous coronary interventions. RESULTS: The average optimal viewing angle for ostial left main stenting was left anterior oblique (LAO) 37°, cranial (CRA) 22° (95% confidence interval [CI]: LAO 33° to 40°, CRA 19° to 25°) and for ostial right coronary stenting was LAO 79°, CRA 41° (95% CI: LAO 74° to 84°, CRA 37° to 45°). Estimated mean optimal viewing angles for bifurcation stenting were as follows: left main: LAO 0°, caudal (CAU) 49° (95% CI: right anterior oblique [RAO] 8° to LAO 8°, CAU 43° to 54°); left anterior descending with first diagonal branch: LAO 11°, CRA 71° (95% CI: RAO 6° to LAO 27°, CRA 66° to 77°); left circumflex bifurcation with first marginal branch: LAO 24°, CAU 33° (95% CI: LAO 15° to 33°, CAU 25° to 41°); and posterior descending artery and posterolateral branch: LAO 44°, CRA 34° (95% CI: LAO 35° to 52°, CRA 27° to 41°). CONCLUSIONS: CTCA can suggest optimal fluoroscopic viewing angles of coronary artery ostia and bifurcations. As the frequency of use of diagnostic CTCA increases in the future, it has the potential to provide additional information for planning and guiding percutaneous coronary intervention procedures.

Recursive multiresolution convolutional neural networks for 3D aortic valve annulus planimetry.

PURPOSE: Transcatheter aortic valve replacement (TAVR) is the standard of care in a large population of patients with severe symptomatic aortic valve stenosis. The sizing of TAVR devices is done from ECG-gated CT angiographic image volumes. The most crucial step of the analysis is the determination of the aortic valve annular plane. In this paper, we present a fully tridimensional recursive multiresolution convolutional neural network (CNN) to infer the location and orientation of the aortic valve annular plane. METHODS: We manually labeled 1007 ECG-gated CT volumes from 94 patients with severe degenerative aortic valve stenosis. The algorithm was implemented and trained using the TensorFlow framework (Google LLC, USA). We performed K-fold cross-validation with K = 9 groups such that CT volumes from a given patient are assigned to only one group. RESULTS: We achieved an average out-of-plane localization error of (0.7 ± 0.6) mm for the training dataset and of (0.9 ± 0.8) mm for the evaluation dataset, which is on par with other published methods and clinically insignificant. The angular orientation error was (3.9 ± 2.3)° for the training dataset and (6.4 ± 4.0)° for the evaluation dataset. For the evaluation dataset, 84.6% of evaluation image volumes had a better than 10° angular error, which is similar to expert-level accuracy. When measured in the inferred annular plane, the relative measurement error was (4.73 ± 5.32)% for the annular area and (2.46 ± 2.94)% for the annular perimeter. CONCLUSIONS: The proposed algorithm is the first application of CNN to aortic valve planimetry and achieves an accuracy on par with proposed automated methods for localization and approaches an expert-level accuracy for orientation. The method relies on no heuristic specific to the aortic valve and may be generalizable to other anatomical features.