Computed Tomography Angiography-Derived Pericoronary Fat Attenuation Index and AI-Based Cardiovascular Risk Score as a Proposed New Pathway in a Rapid Access Chest Pain Clinic: A Single-Centre Retrospective Registry Study
- 1Department of Cardiology, Translational Cardiovascular Research Group, Milton Keynes University Hospital NHS Foundation Trust, Milton Keynes MK6 5LD, UK
- 2Faculty of Medicine and Health Sciences, University of Buckingham, Buckingham MK18 1EG, UK
- 3Department of Radiology, Milton Keynes University Hospital NHS Foundation Trust, Milton Keynes MK6 5LD, UK
Abstract
Background: As part of the National Health Service England new chest pain pathway implementation project, we tested the impact of an artificial intelligence (AI)-based, CariHeart risk score-based new pathway (NP) compared with the current pathway (CP). Methods: One hundred and thirty-five consecutive patients referred to the Rapid Access Chest Pain Clinic who underwent Cardiac Computed Tomography Angiography (CCTA) for the assessment of suspected coronary artery disease (CAD) were included. CCTA images were analysed for qualitative plaque characterisation and grading of coronary artery stenosis and CariHeart risk score (derived from the perivascular fat attenuation index (FAI) score, plaque burden and clinical risk factors) as part of this analysis. We assessed the proportion of patients with non-obstructive CAD classified as high-risk by CariHeart and the differences between the pathways in initiation/intensification and de-escalation of cardiovascular preventive therapy (CVPT; statin, aspirin, ezetimibe and colchicine). CP was based on nurse assessment and CCTA with subsequent cardiologist recommendation and the NP was based on the same in addition to the CariHeart risk assessment-based management. Results: The average age was 61 years, with 47% female and 82% Caucasian. CCTA showed that 83% of patients had non-obstructive CAD (no or <50% stenosis), 16% had moderate CAD (50–69% stenosis), and 1.5% had severe or chronic total occlusion. The CariHeart score indicated low risk (<1% risk) in 3%, intermediate risk (1–5%) in 38%, and high risk (>5%) in 59% of patients. Based on the CariHeart risk score, 54% of patients with non-obstructive CAD were classified as high-risk, whereas 17% of patients with obstructive (≥50%) CAD were classified as intermediate-risk and none as low-risk. Among 133 patients with recommendations under both pathways, escalation of CVPT was recommended for 30% with the CP and 84% with the NP (McNemar p < 0.0001); the rest continued their existing therapy. Under the CP, escalation was more frequent in patients with higher FAI score centiles and higher CariHeart risk; under the NP, escalation rates were high in all categories. No de-escalation of medication was recommended in either of the pathways. In patients with no or mild CAD on the local CCTA report, escalation was recommended more often with the NP than with the CP (86% vs. 25%). Conclusions: Incorporating the AI-based CariHeart risk score into the chest pain pathway classified a high proportion of patients with non-obstructive CAD as high-risk and was associated with a higher rate of recommended escalation of CVPT than the CP. Whether this improves clinical outcomes requires prospective evaluation.
Related
- Comprehensive coronary computed tomography angiography assessment in high-altitude diabetic populations: integrating plaque characteristics, pericoronary inflammation, and coronary function
- Test-bolus peak attenuation for predicting aortic enhancement in coronary computed tomography angiography: an exploratory subgroup analysis
- Incremental diagnostic value of coronary CT angiography-derived pericoronary adipose tissue radiomics for identifying hemodynamically significant coronary stenosis
- Association between lesion-specific fat attenuation index and high-risk plaque on coronary computed tomography angiography
- Post-thrombolysis computed tomography angiography and perfusion features associated with an unfavorable early neurological response: a single-center retrospective study in a clinically selected anterior-circulation cohort
BibTeX
@article{garces2026computed,
title = {{Computed Tomography Angiography-Derived Pericoronary Fat Attenuation Index and AI-Based Cardiovascular Risk Score as a Proposed New Pathway in a Rapid Access Chest Pain Clinic: A Single-Centre Retrospective Registry Study}},
author = {Nerea Sanfeliu Garces and Ahmed Alsinbili and Lucas Costa and Tracy Barnfather and Lynne Cooley and Barbara Kardos and Mike Pashler and Attila Kardos},
journal = {Journal of Clinical Medicine},
year = {2026},
month = nov,
doi = {10.3390/jcm15218639},
url = {https://www.mdpi.com/2077-0383/15/21/8639},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Computed Tomography Angiography-Derived Pericoronary Fat Attenuation Index and AI-Based Cardiovascular Risk Score as a Proposed New Pathway in a Rapid Access Chest Pain Clinic: A Single-Centre Retrospective Registry Study
AU - Nerea Sanfeliu Garces
AU - Ahmed Alsinbili
AU - Lucas Costa
AU - Tracy Barnfather
AU - Lynne Cooley
AU - Barbara Kardos
AU - Mike Pashler
AU - Attila Kardos
PY - 2026
DA - 2026/11/06
JO - Journal of Clinical Medicine
DO - 10.3390/jcm15218639
UR - https://www.mdpi.com/2077-0383/15/21/8639
AB - Background: As part of the National Health Service England new chest pain pathway implementation project, we tested the impact of an artificial intelligence (AI)-based, CariHeart risk score-based new pathway (NP) compared with the current pathway (CP). Methods: One hundred and thirty-five consecutive patients referred to the Rapid Access Chest Pain Clinic who underwent Cardiac Computed Tomography Angiography (CCTA) for the assessment of suspected coronary artery disease (CAD) were included. CCTA images were analysed for qualitative plaque characterisation and grading of coronary artery stenosis and CariHeart risk score (derived from the perivascular fat attenuation index (FAI) score, plaque burden and clinical risk factors) as part of this analysis. We assessed the proportion of patients with non-obstructive CAD classified as high-risk by CariHeart and the differences between the pathways in initiation/intensification and de-escalation of cardiovascular preventive therapy (CVPT; statin, aspirin, ezetimibe and colchicine). CP was based on nurse assessment and CCTA with subsequent cardiologist recommendation and the NP was based on the same in addition to the CariHeart risk assessment-based management. Results: The average age was 61 years, with 47% female and 82% Caucasian. CCTA showed that 83% of patients had non-obstructive CAD (no or <50% stenosis), 16% had moderate CAD (50–69% stenosis), and 1.5% had severe or chronic total occlusion. The CariHeart score indicated low risk (<1% risk) in 3%, intermediate risk (1–5%) in 38%, and high risk (>5%) in 59% of patients. Based on the CariHeart risk score, 54% of patients with non-obstructive CAD were classified as high-risk, whereas 17% of patients with obstructive (≥50%) CAD were classified as intermediate-risk and none as low-risk. Among 133 patients with recommendations under both pathways, escalation of CVPT was recommended for 30% with the CP and 84% with the NP (McNemar p < 0.0001); the rest continued their existing therapy. Under the CP, escalation was more frequent in patients with higher FAI score centiles and higher CariHeart risk; under the NP, escalation rates were high in all categories. No de-escalation of medication was recommended in either of the pathways. In patients with no or mild CAD on the local CCTA report, escalation was recommended more often with the NP than with the CP (86% vs. 25%). Conclusions: Incorporating the AI-based CariHeart risk score into the chest pain pathway classified a high proportion of patients with non-obstructive CAD as high-risk and was associated with a higher rate of recommended escalation of CVPT than the CP. Whether this improves clinical outcomes requires prospective evaluation.
SN - 2077-0383
ER -