High Performance Computing of Personalized Cardio Component Models

Completed

• Conditions: Coronary Artery Disease

• Registration Number: NCT02235883
• Lead Sponsor: Transilvania University of Brasov

Brief Summary

The objective of the HEART Study is to develop a multiscale model of the coronary circulation based on multimodal imaging data (angiography, echocardiography). The predictive comprehensive multi-scale model, developed during this proof-of-concept study, will be used for analyzing functional aspects of coronary artery disease. Model validation is performed against invasive measurements

Detailed Description

The main objective of HEART is the development and validation of a patient-specific multiscale computational model of the coronary circulation with high predictive power in healthy and diseased vessels:

* Comprehensive modeling of the anatomical and hemodynamic phenomena in the coronary circulation based on input data acquired from both angiography and echocardiography;

* Assessment of functional parameters and subsequent validation of the models;

* High performance computing architecture for efficiently addressing the multi-scale complexity, which is a critical requirement for translation into clinical decision making.

Patient measurements used as input data for the multiscale models, or for validating the computational results include:

* Angiography: targeted angiographic views include LAO 30, LAO 30/20 cranial, LAO 30/20 caudal, RAO 30 (RCA views) and RAO 25/25 caudal, RAO 0/40 caudal, RAO 10/40 cranial, LAO 50/20 cranial, LAO 50/20 caudal (LCA views)

* Echocardiography:

* Standard apical acquisitions (2D harmonic 4-chamber, 2-chamber, 3-chamber)

* Color flow Doppler/spectral Doppler (MR, AI, TR evaluation; CW and PW spectral Doppler)

* Three-dimensional imaging (From apical 4-chamber view one ECG-triggered real time complete full volume acquisition including LV and LA - 1 acquisition at low volume per second (VPS) rate and one at high VPS, real time volumetric color flow Doppler, cover MV and AV).

* Non-invasive pressure measurements: cuff-based measurements at the left and right arm, before echocardiography and angiography

* Invasive pressure measurements: rest - aortic pressure (Pa), distal pressure (Pd), Pd/Pa ratio; hyperemia - aortic pressure (Pa), distal pressure (Pd), FFR (Pd/Pa)

Study Timeline

• Study Start: 2014-08
• Study First Submitted: 2014-08-26
• Study First Submitted QC: 2014-09-06
• Study First Posted: 2014-09-10
• Primary Completion: 2016-12
• Study Completion: 2017-01
• Status Verified: 2018-04
• Last Update Submitted: 2018-04-24
• Last Update Posted: 2018-04-25

Recruitment & Eligibility

• Status: COMPLETED
• Sex: All
• Target Recruitment: 94

Inclusion Criteria

• age >18 years
• providing written informed consent
• scheduled to undergo clinically-indicated non-emergent invasive coronary angiography (ICA)
• likelihood of stenosis in one of the main coronary epicardial vessels (LAD, LCx, RCA)

Exclusion Criteria

• prior PCI
• prior bypass surgery
• prior valve surgery or replacement
• significant arrhythmia
• very low systolic BP (<90 mmHg)
• abnormally high HR (>120 bpm)
• significant Left Main disease
• STEMI or NSTEMI
• suspected microvascular disease

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Accuracy of coronary hemodynamic indices extracted from the computational model of the coronary circulation as compared to the invasively measured indices (FFR, basal Pd/Pa)	up to 18 months	Coronary hemodynamic indices (FFR, basal Pd/Pa) are measured invasively and are computed from the computational model of coronary hemodynamics. The diagnostic accuracy of the computed hemodynamic indices is evaluated against the measured hemodynamic indices.

Secondary Outcome Measures

Name	Time	Method
Net reclassification index when using echocardiography data as input parameters for the computational model (in addition to the angiographic data)	up to 18 months	Coronary hemodynamic indices (FFR, basal Pd/Pa) are determined from the computational model in two configurations: with and without echocardiography based input data. We evaluate the net reclassification index of the computed coronary hemodynamic indices when echocardiography based input data is used (as compared to the gold standard represented by the measured coronary hemodynamic indices).
Numerical correlation between computed and measured distal coronary pressure at rest and hyperemia	up to 18 months	Distal coronary pressure at rest and hyperemia are measured invasively and are extracted from the computational model. The correlation between measured and computed quantities is evaluated.

Trial Locations

Locations (1)

Spitalul Clinic de Urgenţǎ Bucureşti; 🇷🇴 Bucharest, Romania