This study will evaluate the ability of device-estimated pulse wave velocity and machine learning methods to improve the prediction of potential symptomatic coronary artery disease
In stable patients with suspected symptomatic coronary artery disease, an estimation of pre-test probability (PTP) and a clinical assessment are used to decide who should be investigated further. PTP has historically been based on age, sex, the nature of chest pain or dyspnea as angina equivalent. It is recommended to continue investigation of all with PTP ≥15%, but also to consider investigation at PTP 5-15% (low-intermediate risk) which is the majority of patients. Despite updates to PTP estimations in the 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes, it has been shown that they overestimate the risk of coronary artery disease. In 2024, the ESC Guidelines were updated to recommend an updated clinical assessment method, the risk factor-weighted clinical likelihood (RF-CL), which is based on symtoms and number of risk factors. It has been shown to have better predictive ability compared to PTP alone, but is still largely based on epidemiological data, which may not be valid for all individuals. Coronary computer tomography angiography (CCTA) is the method becoming increasingly established at low-intermediate risk. An initial, non-invasive strategy with CCTA compared to invasive or more advanced examinations is safe and simple. At the same time, CCTA is resource-intensive, with limited availability, and the examination involves both contrast, radiation and incidental findings. Thus, there is a need to improve the risk estimation. Arterial stiffness assessed by pulse wave velocity is an independent marker for cardiovascular events and has been shown to be independently associated with the degree of coronary artery disease. Arterial stiffness is, however, rarely measured in the clinic as it traditionally has required cumbersome procedures. Newer methods include the brachial single cuff-based Arteriograph and the optical technique photoplethysmography (PPG), widely available in healthcare pulse oximeters, but increasingly also in different consumer devices, often complemented by single-lead ECG. The main aim of this study is to evaluate arterial stiffness and its possible role to improve risk stratification of patients undergoing CCTA for potential coronary artery disease.
Study Type
OBSERVATIONAL
Enrollment
156
Use of standard equipment for usual care
Physiological data acquisition equipment
Use of standard equipment
Danderyd University Hospital
Stockholm, Sweden
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥3
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥3 when adding photoplethysmography (PPG) estimated arterial stiffness to the standard model (including age, sex, symtom score \[0-3\] and number of risk factors \[0-5\]). Coronary artery disease reporting and data system (CAD-RADS) ≥3 refers to the classification of coronary artery disease with at least moderate stenosis as identified on coronary computer tomography angiography. The classification follows the CAD-RADS 2.0 definition. Stenosis is graded in severity from 0-5.
Time frame: Typically within 1 month of enrollment
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥2
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥2 when adding photoplethysmography (PPG) estimated arterial stiffness to models with traditional risk factors for coronary artery disease. Coronary artery disease reporting and data system (CAD-RADS) ≥2 refers to the classification of coronary artery disease with at least mild stenosis as identified on coronary computer tomography angiography. The classification follows the CAD-RADS 2.0 definition. Stenosis is graded in severity from 0-5.
Time frame: Typically within 1 month of enrollment
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥3 by Arterigraph
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥3 when adding Arteriograph-estimated arterial stiffness to the standard model (including age, sex, symtom score \[0-3\] and number of risk factors \[0-5\]). Coronary artery disease reporting and data system (CAD-RADS) ≥3 refers to the classification of coronary artery disease with at least moderate stenosis as identified on coronary computer tomography angiography. The classification follows the CAD-RADS 2.0 definition. Stenosis is graded in severity from 0-5.
Time frame: Typically within 1 month of enrollment
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥2 by Arterigraph
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥2 when adding Arteriograph-estimated arterial stiffness to models with traditional risk factors for coronary artery disease. Coronary artery disease reporting and data system (CAD-RADS) ≥2 refers to the classification of coronary artery disease with at least mild stenosis as identified on coronary computer tomography angiography. The classification follows the CAD-RADS 2.0 definition. Stenosis is graded in severity from 0-5.
Time frame: Typically within 1 month of enrollment
Improved area under receiver operating curve (ROC) to predict Coronary artery calcium score
Improved area under receiver operating curve (ROC) to predict coronary artery calcium (CAC) score when adding photoplethysmography (PPG) or Arteriograph estimated arterial stiffness to models with traditional risk factors for coronary artery disease. PPG-ECG signals used in machine learning and advanced modelling may further improve the prediction. The coronary artery calcium (CAC) score is a measure of the amount of calcified plaque in the coronary arteries, as identified on coronary computer tomography angiography. Higher CAC scores are associated with increased risk of coronary artery disease and future cardiovascular events.
Time frame: Typically within 1 month of enrollment
Number of patients diagnosed with acute or chronic coronary artery disease
As safety outcome; proportion of those who our model estimated as low risk and then diagnosed with acute or chronic coronary artery disease in the year following inclusion in the study.
Time frame: 1 year after enrollment
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥3 by adding ECG
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥2 when adding estimated arterial stiffness and machine-learning interpretation of ECG to the standard model (including age, sex, symtom score \[0-3\] and number of risk factors \[0-5\]).
Time frame: Typically within 1 month of enrollment
Machine learning analysis of photoplethysmography to predict CAD-RADS ≥2
Improved area under receiver operating curve (ROC) to predict CAD-RADS ≥2 when adding machine-learning interpretation of the photoplethysmography (PPG) signal to models based on traditional risk factors for coronary artery disease.
Time frame: Typically within 1 month of enrollment
Machine learning analysis of photoplethysmography to predict aortic stenosis
Machine learning analysis of photoplethysmography (PPG) to predict the presense of aortic stenosis (mild-moderate-severe) on cardiac ultrasound.
Time frame: Typically within 1 month of enrollment
Machine learning analysis of photoplethysmography to predict cardiac function
Machine learning analysis of photoplethysmography (PPG) to predict systolic and diastolic cardiac function assessed by cardiac ultrasound.
Time frame: Typically within 1 month of enrollment