Assessing the diagnostic accuracy of cardisiography in the diagnosis of coronary heart disease: CSG-DR study (A pilot study)

Stress echocardiography and invasive coronary angiography in the current diagnostic algorithm for coronary artery disease (A case of Vladivostok Clinical Hospital No. 1)

Stress echocardiography (Stress-ECHOCG) is a noninvasive method for diagnosing coronary artery disease (CAD) that allows for the detection of transient abnormalities in local contractility during exercise. Objective: To evaluate the role of stress echocardiography in diagnosing stable coronary artery disease (CAD) compared to invasive coronary angiography (ICA). Materials and methods. Stress-ECHOCG was performed on 34 patients with suspected CAD at the Functional Diagnostics Department of Vladivostok Clinical Hospital No 1. The Stress-ECHOCG results were then compared with those of ICA. Results. The sensitivity of stress echocardiography was 96.8%, and the positive predictive value was 90.9%. A positive stress echocardiogram was associated with stenoses in the basins of the anterior descending artery (r = 0.91), the circumflex artery (r = 0.89), and the right coronary artery (r = 0.91). Conclusion. Stress echocardiography plays a crucial role in the diagnostic algorithm for coronary artery disease, offering high accuracy and safety. Using stress echocardiography helps optimize patient selection for invasive testing and reduces the number of unnecessary coronary angiograms.

Diagnostic accuracy of artificial intelligence-enhanced coronary CT angiography for detecting functionally significant coronary artery disease: A systematic review and meta-analysis.

BACKGROUND Coronary computed tomography angiography (CCTA) is widely used to evaluate suspected coronary artery disease (CAD), but its ability to determine the functional significance of coronary stenoses remains limited. Artificial intelligence (AI)-enhanced CCTA has emerged as a promising noninvasive approach to improve ischemia assessment. OBJECTIVES To evaluate the diagnostic accuracy of AI-based CCTA for detecting hemodynamically significant CAD using invasive reference standards. METHODS We conducted a systematic review and diagnostic test accuracy meta-analysis in accordance with PRISMA-DTA guidelines. Studies evaluating AI algorithms applied to CCTA for the detection of functionally significant CAD were eligible if invasive fractional flow reserve (FFR) or invasive coronary angiography served as the reference standard. Risk of bias was assessed using QUADAS-2. The primary analysis was restricted to studies using FFR ≤ 0.80. Diagnostic performance was estimated using a bivariate random-effects hierarchical summary receiver operating characteristic (HSROC) model. Prespecified sensitivity analyses evaluated the effects of alternative reference standards, study quality, verification strategy, AI methodology, and unit of analysis. RESULTS Thirty-five studies involving approximately 8400 participants met the eligibility criteria, of which 18 contributed to the quantitative synthesis. For studies using FFR ≤ 0.80 as the reference standard (13 studies), pooled sensitivity was 0.823 (95% CI, 0.761-0.872) and pooled specificity was 0.820 (95% CI, 0.732-0.883). The bivariate HSROC model produced similar estimates (sensitivity 0.827; specificity 0.820). Sensitivity analyses excluding studies at high risk of bias and including studies using alternative physiological reference standards (iFR ≤ 0.89 or FFR-based composite definitions) demonstrated comparable diagnostic performance. Exploratory subgroup analyses showed generally consistent accuracy across AI methodologies, verification strategies, and units of analysis, although heterogeneity remained. CONCLUSIONS AI-enhanced CCTA demonstrates good and balanced diagnostic accuracy for identifying functionally significant CAD compared with invasive reference standards. Diagnostic performance remained robust across multiple sensitivity analyses, supporting the potential role of AI-assisted CCTA as a noninvasive gatekeeper to invasive coronary angiography. Further prospective multicenter studies using standardized AI algorithms and external validation are needed before widespread clinical implementation.

CAD-RADS 2.0-guided diagnostic pathways in chronic coronary syndromes: a real-world Polish registry of over 10 000 patients.

BACKGROUND Coronary computed tomography angiography (CCTA) reported with Coronary Artery Disease Reporting and Data System (CAD-RADS) 2.0 is increasingly used in chronic coronary syndromes, but real-world adherence to the recommended downstream pathways is poorly characterized. AIMS To describe the population referred for CCTA in a Polish tertiary center, the distribution of CAD-RADS 2.0 categories by sex and age, and adherence to guideline-recommended downstream pathways. METHODS We retrospectively analyzed 10 005 consecutive patients who underwent CCTA between July 1, 2022 and December 31, 2024. Downstream procedures were identified in the National Health Fund registry. A pre-specified multivariable logistic regression model (CAD-RADS category, age, sex) described referral for invasive coronary angiography (ICA); discrimination was quantified as the area under the receiver-operating-characteristic curve (AUC). RESULTS Women constituted 57.8% of the cohort and were older than men (mean 66.4 vs. 64.1 years; P < 0.001). The distribution of categories differed by sex (P < 0.001). Among patients with CAD-RADS 0-2, 18.5% underwent at least one additional test. Among patients with CAD-RADS 3, 41.8% proceeded directly to ICA without prior functional testing and 17.7% followed a functional-testing-first pathway. Among patients with CAD-RADS 4-5, 26.0% did not undergo ICA and 17.8% had no further testing. CAD-RADS category dominated referral for ICA (full model AUC 0.886; 95% confidence interval, 0.877-0.894; CAD-RADS alone 0.883; age and sex alone 0.677). CONCLUSIONS Divergence between observed and guideline-recommended pathways was evident across all CAD-RADS categories, indicating a need for structured implementation of CAD-RADS-guided care pathways.

AI-Assisted Characterization of Coronary High-Risk Plaques in Troponin-Negative Patients Presenting with Atypical Chest Pain: A 256-Slice CCTA Analysis

Background: While coronary computed tomography angiography (CCTA) is an established modality for evaluating chest pain, there is limited up-to-date data on the utility of advanced artificial intelligence (AI)-assisted cardiac software utilizing 256-slice computed tomography (CT) scans to systematically detect and characterize specific high-risk plaque features (HRPF) in the clinical subset of patients with atypical chest pain and troponin-negative (TNEG). Objective: To analyze the characteristics of coronary artery HRPF on CCTA using AI-assisted software and a 256-slice CT scanner in patients presenting with atypical chest pain and TNEG. Materials and Methods: A retrospective analysis was conducted on 118 eligible TNEG patients with documented coronary plaques who underwent CCTA between January 2020 and April 2024. Data were retrieved via the Picture Archiving and Communication System (PACS) and Hospital Information System (HIS). Coronary artery plaque analysis was performed using AI-assisted cardiac software (Cardiac Suite) on a 256-slice CT scanner. The presence and types of HRPF were evaluated, including: 1) low-attenuation plaque (<30 Hounsfield units), 2) positive remodeling, 3) spotty calcification, and 4) the napkin-ring sign. Statistical analyses were performed using the unpaired t-test, chi-square test, and Cramer’s V to determine associations between variables. Results: Of the 118 cases evaluated, obstructive coronary artery disease (CAD) was identified in 66 cases (56.0%) and non-obstructive CAD in 52 cases (44.0%). HRPF were significantly more prevalent in the obstructive CAD group (43/66 cases, 65.2%) compared to the non-obstructive CAD group (5/52 cases, 9.6%). The AI-assisted software efficiently identified and characterized plaque morphology. Low-attenuation plaque was the most common HRPF characteristic, observed in 38 cases (32.2%), predominantly within obstructive lesions (stenosis >50%). A high coronary artery calcium score (CACS >300, percentiles P3 and P4) demonstrated a strong, statistically significant association with HRPF, present in 51 cases (43.2%). Based on these data, it is hypothesized that the transition from P2 (CACS 101-300) to P3/P4 (CACS >300) corresponds to a biological shift from non-HRPF to HRPF. Notably, a 36-year-old male smoker presented with distinct obstructive HRPF despite a CACS of 0. Conclusion: AI-assisted cardiac software provides rapid, precise identification and characterization of coronary plaques, particularly vulnerable HRPF. Upgrading CT infrastructure with integrated AI-assisted analysis tools significantly enhances diagnostic accuracy. This software acts as a crucial clinical decision-making aid to accurately rule out or confirm coronary etiologies in patients presenting with atypical chest pain and TNEG.

Diagnostic Performance of Coronary CT Angiography versus Invasive Coronary Angiography in Patients Undergoing Percutaneous Coronary Intervention: A Cross-Sectional Study.

Standard clinical and stress electrocardiogram variables to exclude left main and left main-equivalent disease: the MASTER study.

BACKGROUND AND AIMS A simple diagnostic method able to reliably exclude left main (LM) coronary artery disease (CAD) or LMCAD-equivalent would expand implementation of an initial non-invasive strategy in patients with chronic coronary syndrome (CCS). This study assessed the diagnostic utility of an approach using clinical and ECG stress testing (EST) variables in excluding LMCAD/LMCAD-equivalent in CCS patients. METHODS In a multicentre case-control study, CCS patients undergoing invasive coronary angiography (CAG) after a maximal EST were evaluated. Cases were patients with angiographic ≥ 50% LM stenosis or ≥70% stenosis of both proximal left anterior descending and proximal circumflex arteries, matched with similar patients without them (controls) in a 1:3 ratio. A risk model developed through logistic regression was internally and externally validated. RESULTS Three hundred and thirty-five cases were matched with 797 controls. The model area under the curve (AUC) was .78. Assuming LMCAD prevalence of 5% and a misclassification cost ratio of 1:100 (ratio of cost of performing CAG in a control to cost of not performing CAG in a case), negative predictive value was 98.2%. Thus, CAG could be safely avoided in 41% of patients, missing one LMCAD/LMCAD-equivalent diagnosis for every 58 CAGs safely spared in patients without them. CONCLUSIONS Among CCS patients, LMCAD/LMCAD-equivalent can be excluded with high negative predictive value through a model based on clinical and EST parameters, allowing initial non-invasive management of most patients able to exercise. This approach is potentially useful particularly in communities where access to computed tomography coronary angiography is limited.