Author

Serban-Georges Puricel

View on Semantic Scholar

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Open access Jul 2026

Specificity of the Occlusion Myocardial Infarction-Trained PMcardio “Queen of Hearts” AI-ECG in Inflammatory Myopericardial Syndromes

Background Inflammatory myopericardial syndromes (IMPS)—including pericarditis, myocarditis, and overlap phenotypes—frequently present with chest pain and electrocardiographic abnormalities that can mimic occlusion myocardial infarction (OMI). High electrocardiographic specificity is essential to avoid unnecessary catheterization laboratory activation. Objectives To evaluate the ECG-level specificity of the OMI-trained PMcardio Queen of Hearts (QoH) AI-ECG model for OMI classification in patients with IMPS. Methods We conducted a single-centre retrospective study including consecutive emergency department patients with adjudicated IMPS (2010-2025) at the University & Hospital of Fribourg. Obstructive coronary artery disease was excluded by coronary angiography or cardiac magnetic resonance imaging according to contemporary guidelines. Index ECGs were analysed using QoH. Specificity was compared with European Society of Cardiology (ESC) STEMI ECG criteria, the 4-variable formula (when applicable), and 3 large language models (LLMs; ChatGPT 5.2, Claude Opus 4.5, and Gemini 3). The primary endpoint was ECG-level specificity. Because only adjudicated IMPS cases were included, the analysis was designed to estimate ECG-level specificity rather than overall diagnostic accuracy. Results Among 242 IMPS patients, QoH correctly classified 230 as non-OMI (specificity: 95.0%; 95% CI: 91.5-97.1). ESC STEMI ECG criteria showed 78.1% specificity, and the 4-variable formula demonstrated 73.3% specificity in anterior presentations (n = 30). All LLMs exhibited substantially lower specificity (26.9% to 70.2%; all P < 0.001 vs QoH). Discordant ESC-positive/QoH-negative cases (n = 43) predominantly showed diffuse, nonterritorial inflammatory ECG patterns. Specificity was higher in (myo)pericarditis than in (peri)myocarditis (97.7% vs 87.7%, P = 0.004). Conclusions In IMPS, the OMI-trained QoH model demonstrated high specificity and fewer false-positive OMI classifications than guideline ECG-STEMI criteria, ECG-based scores, and general-purpose LLMs.

W. Bennar, Dorian Garin, R. M. Saidi et al. · JACC: Advances · 1 citation

Review Aug 2026

Large language models in interventional cardiology: current evidence and future directions.

Large language models (LLMs) have rapidly emerged as a transformative class of artificial intelligence systems capable of understanding and generating human-like text from vast corpora of clinical and scientific literature. While their use in general cardiology has been extensively reviewed, their specific role in interventional cardiology and within the catheterization laboratory (cath lab) environment remains less well characterized. This narrative review synthesizes the current evidence on LLM applications across the interventional cardiology workflow, including coronary revascularization decision-making, multidisciplinary Heart Team support, structural heart intervention planning, periprocedural communication, and acute cath lab decision support. Recent studies suggest that contemporary LLMs such as ChatGPT-4 (OpenAI), Claude (Anthropic), and Gemini (Alphabet, Inc.) can achieve clinically meaningful concordance with expert Heart Team recommendations for percutaneous coronary intervention vs coronary artery bypass grafting, and that their outputs may approach or even match those of early-career interventional cardiologists in simulated emergency scenarios. However, important limitations persist, including variable accuracy across prompt formats, susceptibility to hallucinations, lack of multimodal integration with angiographic and intravascular imaging, opacity of training sources, and unresolved medicolegal issues. The authors discuss the current evidence base, highlight promising avenues such as multimodal LLMs and retrieval-augmented generation tools that leverage current guidelines, and outline regulatory and ethical considerations that must accompany clinical adoption. LLMs are unlikely to replace interventional cardiologists in the foreseeable future, but they may meaningfully support decision-making, education, and patient communication in a domain where speed, complexity, and risk converge.

Ioannis Skalidis, Lisa Simioni, G. Beretta et al. · The Journal of invasive card... · 0 citations