Background/Objectives: Large language models (LLMs), such as ChatGPT, have emerged as potential clinical support tools to enhance precision in personalized patient care, but their reliability in radiological image interpretation remains uncertain. The primary aim of our study was to evaluate the diagnostic accuracy of ChatGPT-4o in interpreting chest X-rays (CXRs) and abdominal X-rays (AXRs) by comparing its performance to expert radiology findings, whilst secondary aims were diagnostic confidence and patient safety. Methods: A total of 500 X-rays, including 257 CXR (51.4%) and 243 AXR (48.5%), were analyzed. Diagnoses made by ChatGPT-4o were compared to expert interpretations. Confidence scores (1–4) were assigned and responses were evaluated for patient safety. Results: ChatGPT-4o correctly identified 345 of 500 (69%) pathologies (95% CI: 64.81–72.9). For AXRs 175 of 243 (72.02%) pathologies were correctly diagnosed (95% CI: 66.06–77.28), while for CXRs 170 of 257 (66.15%) were accurate (95% CI: 60.16–71.66). The highest detection rates among CXRs were observed for pulmonary edema, tumor, pneumonia, pleural effusion, cardiomegaly, and emphysema, and lower rates were observed for pneumothorax, rib fractures, and enlarged mediastinum. AXR performance was highest for intestinal obstruction and foreign bodies, and weaker for pneumoperitoneum, renal calculi, and diverticulitis. Confidence scores were higher for AXRs (mean 3.45 ± 1.1) than CXRs (mean 2.48 ± 1.45). All responses (100%) were considered to be safe for the patient. Interobserver agreement was high (kappa = 0.920), and reliability (second prompt) was moderate (kappa = 0.750). Conclusions: ChatGPT-4o demonstrated moderate accuracy for the interpretation of X-rays, being higher for AXRs compared to CXRs. Improvements are required for its use as efficient clinical support tool.

The severity and extent of whole heart coronary plaque volume and stenosis can be reliably measured by artificial intelligence-guided quantitative coronary computed tomography angiography (AI-QCT). Limited data are available on the potential incremental prognostic value compared with currently recommended qualitative coronary computed tomography angiography (CTA) reads and the coronary artery calcium score (CACS). The aim of this study was to evaluate the prognostic value of AI-QCT compared with human coronary CTA reads, including the CAD-RADS (Coronary Artery Disease-Reporting and Data System), CACS, and the modified Duke Index. CONFIRM2 (Quantitative COroNary CT Angiography Evaluation For Evaluation of Clinical Outcomes: An InteRnational, Multicenter Registry) is a multicenter, international, observational cohort study of patients undergoing clinically indicated coronary CTA with follow-up for major adverse cardiac events (MACE). Asymptomatic patients and those with cardiac history were excluded. Coronary artery disease presence, extent, and composition were quantified by AI-QCT across the coronary tree, yielding 24 patient-, vessel-, and plaque-level variables. On the basis of prior analyses, noncalcified plaque burden and diameter stenosis were identified as the strongest predictors and combined for statistical modeling as "AI-QCT." Comparator computed tomography scores included CAD-RADS, CACS, and the modified Duke Index, whereas clinical predictors were summarized in the risk factor-weighted clinical likelihood score. Area under the curve (AUC) and continuous net reclassification index (NRI) were calculated to assess the incremental value. The primary endpoint was MACE (death, myocardial infarction [MI], stroke, heart failure, late revascularization, or hospital stay for unstable angina), and the secondary endpoint was death or MI. In 1,916 patients with all risk scores available, 87 (4.5%) MACE and 27 (1.4%) death/MI events occurred during 3 years of follow-up. There was a stepwise risk increase with higher coronary artery disease classifications with CAD-RADS and CACS. The addition of AI-QCT significantly improved risk stratification for MACE compared with CAD-RADS (AUC: 0.81 vs 0.79; P < 0.001 and NRI: 0.47; P < 0.001), CACS (AUC: 0.79 vs 0.70; P < 0.001 and NRI 0.61; P < 0.001), the modified Duke Index (AUC: 0.81 vs 0.76; P < 0.001 and NRI: 0.52; P < 0.001), and CAD-RADS + CACS model (AUC: 0.81 vs 0.79; P = 0.004 and NRI: 0.54; P < 0.001). AI-QCT also improved discrimination when results were adjusted for the risk factor-weighted clinical likelihood and for the prediction of death/MI. Excluding 195 patients with severe stenosis (≥70%), in a multivariable model of CAD-RADS and AI-QCT, only AI-QCT was significantly associated with MACE and death/MI, and AI-QCT significantly improved risk stratification compared with CAD-RADS for MACE (AUC: 0.77 vs 0.72; P < 0.001 and NRI: 0.54; P < 0.001) and death/MI (AUC: 0.81 vs 0.73; P = 0.011 and NRI: 0.69; P = 0.001). AI-QCT provided incremental prognostic information compared with CAD-RADS 2.0, CACS, and the modified Duke Index for the prediction of MACE as well as the secondary endpoint of death or nonfatal MI.

Photon-counting detector computed tomography (PCD-CT) represents a revolutionary new generation of computed tomography (CT) for the imaging of patients with cardiovascular diseases. Since its commercial market introduction in 2021, numerous studies have identified advantages of this new technology in the field of cardiovascular imaging, including improved image quality due to an enhanced contrast-to-noise ratio, superior spatial resolution, reduced artifacts, and a reduced radiation dose. The aim of this narrative review was to discuss the current scientific literature, and to find answers to the question of whether PCD-CT has yet led to a true step-change and significant progress in cardiovascular imaging.

Plaque assessment by quantitative coronary CT angiography has demonstrated to correlate highly with intravascular ultrasound and optical coherence tomography, and these modalities have shown strong prognostic value. The purpose of this study was to identify the prognostic value of artificial intelligence-guided quantitative CCTA (AI-QCT) for major adverse cardiovascular events (MACE) against the risk factor-weighted clinical likelihood model. The CONFIRM2 (COroNary CT Angiography Evaluation For Evaluation of Clinical Outcomes: An InteRnational, Multicenter Registry) is a multicenter, international, observational cohort study that included patients with clinically indicated CCTA and follow-up for MACE. Patients without cardiac symptoms and prior coronary artery disease (CAD) were excluded. Across the entire coronary artery tree, the presence, extent, and composition of CAD were analyzed by an AI-QCT software, and 24 variables at a patient, vessel, and plaque level were derived, including percent luminal narrowing, remodeling index, plaque volumes (total, calcified, noncalcified, low attenuation), and plaque composition. The primary MACE endpoint was defined as a composite of all-cause death, myocardial infarction (MI), stroke, congestive heart failure, late revascularizations, and hospitalization for unstable angina. The secondary MACE endpoint was defined as all-cause death and MI. A total of 3,551 patients (age 58.8 ± 12.5 years, 50.5% male) were followed for a median of 4.27 (IQR: 3.47-5.08) years during which 167 (4.7%) events occurred. After excluding collinear variables, diameter stenosis (HR: 1.25 [95% CI: 1.18-1.32]) per 10% increase and noncalcified plaque volume (HR: 1.07 [95% CI: 1.03-1.11]) per 50 mm³ increase were the only independent predictors for MACE. In multivariable modeling, the discriminatory value defined by area under the curve (AUC) improved from 0.63 (95% CI: 0.58-0.67) based on the risk factor-weighted clinical likelihood model to 0.76 (95% CI: 0.77-0.80), P < 0.001 when adding AI-QCT-based diameter stenosis and noncalcified plaque volume. A similar improvement in risk prediction was seen when adding AI-QCT (AUC 0.77; P < 0.001) to a model with traditional risk factors, age, and sex (AUC: 0.67). In addition, AI-QCT significantly improved discrimination compared to the atherosclerotic cardiovascular disease risk score (AUC: 0.63; 95% CI: 0.58-0.68) to 0.75 (95% CI: 0.69-0.80; P < 0.001). Similar results were seen for the secondary MACE endpoint of death/MI. This first multicenter global registry with AI-guided quantitative CT identified noncalcified plaque burden and increment in stenosis severity as the most powerful predictors of MACE, demonstrating the interplay between traditional and novel measures of the severity of CAD. Standardized and rapid quantitative assessment of CAD may improve clinical implementation of multidimensional assessment of CAD as a cornerstone for risk assessment.

Background/Objective: Left atrial (LA) fibrosis imaging improves the guidance of LA catheter ablation. Cardiac computed tomography (CT) may be a reasonable alternative to CMR. The aim was to evaluate late enhancement (LE) fibrosis mapping by CT, and to correlate the results with low-voltage areas on electroanatomical mapping (EAM). Methods: In patients with atrial fibrillation who underwent 128-slice dual-source CT angiography (CTA) prior to LA catheter ablation, an additional LE-CT scan was performed 7 min after CTA. (1) Left atrial wall thickness (LAWT) was measured at three sites along the LA ridge. (2) Late enhancement (LE) was quantified co-axially aligned to LAWT and compared with low-voltage areas (LVA) on EAM. Results: Of 137 patients (age: 59.8 years; 27.7% females), 108 were included. The prevalence of LE was higher in patients with LAWT > 2 mm compared with 1.5 mm, with 78 (91.7%) vs. 77 (80.2%) (p = 0.022). Of 78 patients with LE, 60 (77.1%) had focal, 13 (16.5%) had diffuse, and 5 (6.3%) had mixed LE patterns. The CT density of focal LE was not different from that of diffuse patterns (104.2 +/− 21 HU vs. 98.9 +/− 18 HU; p = 0.360). Increasing LAWT and LE-HU were weakly correlated (r = 0.229; p = 0.041). LA wall artifacts had higher CT density compared with LE (154.1 HU vs. 114.2 HU; p = 0.002). The effective radiation dose was 0.95 mSv (range, 0.52–1.2 mSv) for LE-CT. The agreement of LE-CT was 80% for LVA < 0.5 mV and 86.6% for LVA < 0.7 mV in a subset of 30 patients. Conclusions: Left atrial fibrosis mapping by LE-CT is feasible. Late enhancement was found more frequently in LAWTs of more than 2 mm, and LE was correlated with increasing LA remodeling and low-voltage areas.

High exercise volumes may have deleterious effects on the cardiovascular system, and the upper thresholds for "safe" volumes are unclear. To evaluate if high-exercise volumes influence the coronary artery disease (CAD) profile by coronary CTA, cardiovascular outcomes, compared with traditional CVRF. 802 patients (age 57.6 ± 10.95 years;38.1 % women) undergoing coronary CTA were enrolled. Self-reported exercise habits were collected: Type, volume (frequency/week and time/per unit); and years of training. Endpoints were: CTA: coronary stenosis severity (CADRADS); high-risk-plaque (HRP); coronary artery calcium score (CAC), and MACE. 478 subjects were included. 100 with high-(H)-exercise level (>3–5x/week and 1–3 h per unit) were propensity-score matched with 124 sedentary controls. The CVRF dyslipidemia (p = 0.393) and age were similar in both groups (p = 0.328), while arterial hypertension (p = 0.016), diabetes (p = 0.032), and the BMI (p < 0.001) were lower in athletes. CAC (80.5 vs 107.7 AU: p = 0.820, CADRADS: p = 0.394) and MACE-rates were not different. Follow-up was 3.95±1.0 years. After matching for sex, HRP was 1.6-fold less frequent in the H-group (17 % vs 32.8 %; p = 0.231; OR 1.58 (95 % CI: 0.787–3.222) and smoking was more prevalent in controls. There was no difference in HRP after adjusting for all CVRF. MET was mean 8.78 ± 3.5; weekly training volume 9 h. The majority (78 %) reported >5 -10 years of training. High-exercise levels (up to 9h/week) are safe - without a deleterious impact on the coronary atherosclerosis profile.