Agreement of an invasive angiography-derived computational model of coronary microvascular physiology against continuous thermodilution

Coronary microvascular dysfunction is associated with myocardial ischaemia and an adverse prognosis, but accurate invasive assessment is technically challenging, requires dedicated hardware and increases procedure time and cost. Assessment can be performed with three-dimensional (3D) computational fluid dynamics (CFD) modelling, but this technique lacks robust in vivo validation. In this study, we compared the accuracy of a 3D CFD method against continuous thermodilution (CIT). Patients with acute and chronic coronary syndromes, undergoing CIT and angiography-derived assessment, were prospectively recruited from two tertiary cardiac centres. Invasive coronary angiographic images were used to reconstruct 3D anatomy in silico. Microvascular resistance reserve (MRR) and absolute flow were computed using 3D CFD in reconstructed coronary arteries. Invasive pressure measurements, recorded during fractional flow reserve assessment, informed the CFD boundary conditions. Paired flow results were available for 131 arteries from 89 patients. Median FFR was 0.81 [0.70 - 0.89] and 108 (81%) vessels contained an epicardial stenosis. Median computed MRR was 2.31 [1.83 - 3.00], which was significantly lower than CIT assessed MRR (2.79 [2.21 - 3.52], z = 3.57, p = 0.0004). There was evidence of a moderate relationship between computed and measured MRR (ρ = 0.58, p < 0.0001) and area under the receiver operator characteristic curve was 0.77 (95% CI 0.68-0.86), indicating fair classification. There was a moderate relationship between computed and measured hyperaemic vessel inlet flow (ρ = 0.44, p < 0.0001). For both MRR and flow, agreement improved at lower, more clinically relevant, values. In this clinical validation study, when compared with CIT, the novel CFD method demonstrated a moderate correlation, and fair diagnostic accuracy. This CFD method may be a useful, low-cost and less-invasive tool in the assessment of microvascular physiology that could complement current invasive techniques.

Authors

Institutions

Publication Details

Journal
PLOS Digital Health
Published
2026-11-13
DOI
https://doi.org/10.1371/journal.pdig.0001830
Citations
1
Primary Topic
Coronary Interventions and Diagnostics
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Agreement of an invasive angiography-derived computational model of coronary microvascular physiology against continuous thermodilution

Oxami Study Investigators, Krzysztof Czechowicz, Julian P. Gunn, Federico Marin et al.
1 citations
PLOS Digital Health
Coronary Interventions and Diagnostics
article

Agreement of an invasive angiography-derived computational model of coronary microvascular physiology against continuous thermodilution

Oxami Study Investigators, Krzysztof Czechowicz, Julian P. Gunn, Federico Marin, Eron Yones, Thomas R. Keeble, Giovanni Luigi De Maria, Alessandro Maino, Dominika Ciupek, Harry Saxton, Tom Newman, Rebecca C. Gosling, Alberto Biancardi, Andrew James Narracott, Karolina Tlałka, Xu Xu, Paul Morris, Samer Fawaz, Daniel James Taylor, Kenneth Anigboro
article en
1 citations

Abstract

Coronary microvascular dysfunction is associated with myocardial ischaemia and an adverse prognosis, but accurate invasive assessment is technically challenging, requires dedicated hardware and increases procedure time and cost. Assessment can be performed with three-dimensional (3D) computational fluid dynamics (CFD) modelling, but this technique lacks robust in vivo validation. In this study, we compared the accuracy of a 3D CFD method against continuous thermodilution (CIT). Patients with acute and chronic coronary syndromes, undergoing CIT and angiography-derived assessment, were prospectively recruited from two tertiary cardiac centres. Invasive coronary angiographic images were used to reconstruct 3D anatomy in silico. Microvascular resistance reserve (MRR) and absolute flow were computed using 3D CFD in reconstructed coronary arteries. Invasive pressure measurements, recorded during fractional flow reserve assessment, informed the CFD boundary conditions. Paired flow results were available for 131 arteries from 89 patients. Median FFR was 0.81 [0.70 - 0.89] and 108 (81%) vessels contained an epicardial stenosis. Median computed MRR was 2.31 [1.83 - 3.00], which was significantly lower than CIT assessed MRR (2.79 [2.21 - 3.52], z = 3.57, p = 0.0004). There was evidence of a moderate relationship between computed and measured MRR (ρ = 0.58, p < 0.0001) and area under the receiver operator characteristic curve was 0.77 (95% CI 0.68-0.86), indicating fair classification. There was a moderate relationship between computed and measured hyperaemic vessel inlet flow (ρ = 0.44, p < 0.0001). For both MRR and flow, agreement improved at lower, more clinically relevant, values. In this clinical validation study, when compared with CIT, the novel CFD method demonstrated a moderate correlation, and fair diagnostic accuracy. This CFD method may be a useful, low-cost and less-invasive tool in the assessment of microvascular physiology that could complement current invasive techniques.

PLOS Digital HealthVol. 5(11)
Essex Cardiothoracic Centre (GB), Insigneo (GB), Sano – Centre for Computational Personalised Medicine International Research Foundation (PL), University of Sheffield (GB)
National Institute for Health and Care Research, Department of Health and Social Care, Engineering and Physical Sciences Research Council, NIHR Sheffield Biomedical Research Centre
Good health and well-being
Coronary Interventions and Diagnostics
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon