Stephane Fournier

Dr. Stephane Fournier


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 April 2023 in European Heart Journal - Digital Health
Verified
Journal Article
Open access
PDF file
ISIoannis Skalidis
Ioannis Skalidis
ACAurelien Cagnina
Aurelien Cagnina
WLWongsakorn Luangphiphat
Wongsakorn Luangphiphat
TMThabo Mahendiran
Thabo Mahendiran
OMOlivier Muller
Olivier Muller
SFStephane Fournier
Stephane Fournier
Chat Generative Pre-trained Transformer (ChatGPT) is currently a trending topic worldwide triggering extensive debate about its predictive power, its potential uses, and its wider implications. Recent publications have demonstrated that ChatGPT can correctly answer questions from undergraduate exams such as the United States Medical Licensing Examination. We challenged it to answer questions from a more demanding, post-graduate exam—the European Exam in Core Cardiology (EECC), the final exam for the completion of specialty training in Cardiology in many countries. Our results demonstrate that ChatGPT succeeds in the EECC.

20 August 2024 in Circulation
Verified
Journal Article
Open access
CCCarlos Collet
Carlos Collet
DMDaniel Munhoz
Daniel Munhoz
TMTakuya Mizukami
Takuya Mizukami
JSjeroen sonck
jeroen sonck
HMHitoshi Matsuo
Hitoshi Matsuo
NJNils P. Johnson
Nils P. Johnson
BACKGROUND: Diffuse coronary artery disease affects the safety and efficacy of percutaneous coronary intervention (PCI). Pathophysiologic coronary artery disease patterns can be quantified using fractional flow reserve (FFR) pullbacks incorporating the pullback pressure gradient (PPG) calculation. This study aimed to establish the capacity of PPG to predict optimal revascularization and procedural outcomes. METHODS: This prospective, investigator-initiated, single-arm, multicenter study enrolled patients with at least one epicardial lesion with an FFR ≤0.80 scheduled for PCI. Manual FFR pullbacks were used to calculate PPG. The primary outcome of optimal revascularization was defined as an FFR ≥0.88 after PCI. RESULTS: A total of 993 patients with 1044 vessels were included. The mean FFR was 0.68±0.12, PPG 0.62±0.17, and the post-PCI FFR was 0.87±0.07. PPG was significantly correlated with the change in FFR after PCI (r=0.65 [95% CI, 0.61–0.69]; P <0.001) and demonstrated excellent predictive capacity for optimal revascularization (area under the receiver operating characteristic curve, 0.82 [95% CI, 0.79–0.84]; P <0.001). FFR alone did not predict revascularization outcomes (area under the receiver operating characteristic curve, 0.54 [95% CI, 0.50–0.57]). PPG influenced treatment decisions in 14% of patients, redirecting them from PCI to alternative treatment modalities. Periprocedural myocardial infarction occurred more frequently in patients with low PPG (<0.62) compared with those with focal disease (odds ratio, 1.71 [95% CI, 1.00–2.97]). CONCLUSIONS: Pathophysiologic coronary artery disease patterns distinctly affect the safety and effectiveness of PCI. PPG showed an excellent predictive capacity for optimal revascularization and demonstrated added value compared with an FFR measurement. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT04789317.

4 September 2025 in Medicina
Verified
Journal Article
Open access
PDF file
ISIoannis Skalidis
Ioannis Skalidis
NMNiccolò Maurizi
Niccolò Maurizi
ASAdil Salihu
Adil Salihu
SFStephane Fournier
Stephane Fournier
SCStéphane Cook
Stéphane Cook
PAPanagiotis Antiochos
Panagiotis Antiochos
Background: Hypertension remains the leading global risk factor for cardiovascular morbidity and mortality, with suboptimal control rates despite guideline-directed therapies. Digital health and artificial intelligence (AI) technologies offer novel approaches for improving diagnosis, monitoring, and individualized treatment of hypertension. Objectives: To critically review the current landscape of AI-enabled digital tools for hypertension management, including emerging applications, implementation challenges, and future directions. Methods: A narrative review of recent PubMed-indexed studies (2019–2024) was conducted, focusing on clinical applications of AI and digital health technologies in hypertension. Emphasis was placed on real-world deployment, algorithmic explainability, digital biomarkers, and ethical/regulatory frameworks. Priority was given to high-quality randomized trials, systematic reviews, and expert consensus statements. Results: AI-supported platforms—including remote blood pressure monitoring, machine learning titration algorithms, and digital twins—have demonstrated early promise in improving hypertension control. Explainable AI (XAI) is critical for clinician trust and integration into decision-making. Equity-focused design and regulatory oversight are essential to prevent exacerbation of health disparities. Emerging implementation strategies, such as federated learning and co-design frameworks, may enhance scalability and generalizability across diverse care settings. Conclusions: AI-guided titration and digital twin approaches appear most promising for reducing therapeutic inertia, whereas cuffless blood pressure monitoring remains the least mature. Future work should prioritize pragmatic trials with equity and cost-effectiveness endpoints, supported by safeguards against bias, accountability gaps, and privacy risks.

3 January 2023 in Open Heart
Verified
Journal Article
Open access
PDF file
TMThabo Mahendiran
Thabo Mahendiran
DTDorina Thanou
Dorina Thanou
OSOrtal Senouf
Ortal Senouf
DMDavid Meier
David Meier
NDNicolas Dayer
Nicolas Dayer
SFStephane Fournier
Stephane Fournier
Background Angiographic parameters can facilitate the risk stratification of coronary lesions but remain insufficient in the prediction of future myocardial infarction (MI). AIMS We compared the ability of humans, angiographic parameters and deep learning (DL) to predict the lesion that would be responsible for a future MI in a population of patients with non-significant CAD at baseline. Methods We retrospectively included patients who underwent invasive coronary angiography (ICA) for MI, in whom a previous angiogram had been performed within 5 years. The ability of human visual assessment, diameter stenosis, area stenosis, quantitative flow ratio (QFR) and DL to predict the future culprit lesion (FCL) was compared. Results In total, 746 cropped ICA images of FCL and non-culprit lesions (NCL) were analysed. Predictive models for each modality were developed in a training set before validation in a test set. DL exhibited the best predictive performance with an area under the curve of 0.81, compared with diameter stenosis (0.62, p=0.04), area stenosis (0.58, p=0.05) and QFR (0.67, p=0.13). DL exhibited a significant net reclassification improvement (NRI) compared with area stenosis (0.75, p=0.03) and QFR (0.95, p=0.01), and a positive nonsignificant NRI when compared with diameter stenosis. Among all models, DL demonstrated the highest accuracy (0.78) followed by QFR (0.70) and area stenosis (0.68). Predictions based on human visual assessment and diameter stenosis had the lowest accuracy (0.58). Conclusion In this feasibility study, DL outperformed human visual assessment and established angiographic parameters in the prediction of FCLs. Larger studies are now required to confirm this finding.

4 February 2021 in Journal of Cardiovascular Translational Research
Verified
Journal Article
Open access
PDF file
DMDavid Meier
David Meier
ADArnaud Depierre
Arnaud Depierre
ATAntoine Topolsky
Antoine Topolsky
CRChristan Roguelov
Christan Roguelov
MDMarion Dupré
Marion Dupré
SFStephane Fournier
Stephane Fournier
Background: Computed tomography angiography (CTA) is used to plan TAVI procedures. We investigated the performance of pre-TAVI CTA for excluding coronary artery disease (CAD). Methods: In total 127 patients were included. CTA images were analyzed for the presence of ≥ 50% (significant CAD) and ≥ 70% (severe CAD) diameter stenoses in proximal coronary arteries. Results were compared with invasive coronary angiography (ICA) at vessel and patient levels. Primary endpoint was the negative predictive value (NPV) of CTA for the presence of CAD. Results: A total of 342 vessels were analyzable. NPV of CTA was 97.5% for significant CAD and 96.3% for severe CAD. Positive predictive value and accuracy were 44.8% and 87.1% for significant CAD and 56.3% and 94.4% for severe CAD. At patient level, NPV for significant CAD was 88.6%. Conclusion: Pre-TAVI CTA shows good performance for ruling out CAD and could be used as a gatekeeper for ICA in selected patients. Graphical

17 March 2023 in European Journal of Preventive Cardiology
Verified
Journal Article
Open access
PDF file
TMThabo Mahendiran
Thabo Mahendiran
AHAndré Hoepli
André Hoepli
FFFabienne Foster-Witassek
Fabienne Foster-Witassek
HRHans Rickli
Hans Rickli
MRMarco Roffi
Marco Roffi
SFStephane Fournier
Stephane Fournier
Modifiable cardiovascular risk factors (RFs) play a key role in the development of coronary artery disease. We evaluated 20-year trends in RF prevalence among young adults hospitalised with acute coronary syndromes (ACS) in Switzerland. Data were analysed from the Acute Myocardial Infarction in Switzerland (AMIS) Plus registry from 2000 to 2019. Young patients were defined as those aged <50 years. Among 58’028 ACS admissions, 7’073 (14.1%) were young (median 45.6 years, IQR 42.0-48.0), of which 91.6% had at least one modifiable RF, and 59.0% had at least two RFs. Smoking was the most prevalent RF (71.4%), followed by dyslipidaemia (57.3%), hypertension (35.9%), obesity (21.7%) and diabetes (10.1%). Compared with older patients, young patients were more likely to be obese (21.7% vs 17.4%, p < 0.001) and active smokers (71.4% vs 33.9%, p < 0.001). Among young patients, between 2000 and 2019 there was a significant increase in the prevalence of hypertension from 29.0% to 51.3% and obesity from 21.2% to 27.1% (both ptrend < 0.001), but a significant decrease in active smoking from 72.5% to 62.5% (ptrend = 0.02). There were no significant changes in the prevalence of diabetes (ptrend = 0.32) or dyslipidaemia (ptrend = 0.067). Young ACS patients in Switzerland exhibit a high prevalence of RFs, and are more likely than older patients to be obese and smokers. Between 2000 and 2019, RF prevalence either increased or remained stable, except for smoking which decreased but still affected ∼2/3 of young patients in 2019. Public health initiatives targeting RFs in young adults in Switzerland are warranted. We evaluated the prevalence of risk factors (RFs) among young patients admitted with ACS in Switzerland between 2000 and 2019. Young ACS patients in Switzerland exhibited a high prevalence of RFs. There was a significant increase in the prevalence of hypertension and obesity. Despite a significant decrease, active smoking remained the most prevalent RF. These findings strongly suggest that public health initiatives targeting RFs in young adults in Switzerland are warranted.