Wei Yu

Background/Objectives: This study aimed to investigate the associations between preoperative carotid high-resolution magnetic resonance imaging (HR-MRI) plaque features and coronary atherosclerotic burden, and to compare the cross-vascular association patterns of plaque burden-related measures and individual vulnerable plaque features. Methods: This retrospective study included consecutive patients who underwent carotid revascularization and had both preoperative carotid HR-MRI and coronary computed tomography angiography (CTA) between April 2017 and December 2024. Carotid plaque characteristics included calcification, lipid-rich necrotic core (LRNC), intraplaque hemorrhage (IPH), loose matrix (LM), ulceration, maximum wall thickness (MWT), maximum wall area (MWA), and normalized wall index (NWI). Coronary atherosclerotic burden was characterized using coronary artery calcium score (CACS), maximum luminal stenosis, and the number of diseased coronary arteries; the number of coronary arteries with ≥50% stenosis was additionally recorded. Associations between carotid plaque characteristics and coronary atherosclerotic burden were evaluated using correlation analyses and multivariable logistic regression. Results: A total of 162 patients were included (mean age, 66.11 ± 7.54 years; 138 [85.2%] men). CACS ≥ 400 was present in 76 patients (46.9%), obstructive coronary disease (≥50% stenosis in at least one coronary artery) in 96 patients (59.3%), and involvement of ≥3 diseased coronary arteries in 128 patients (79.0%). Greater coronary atherosclerotic burden was consistently associated with higher carotid calcification volume and MWT and NWI (all p < 0.05). After adjustment for clinical risk factors and correction for multiple testing, calcification volume, MWT and NWI remained associated with CACS ≥ 400, obstructive coronary disease, and involvement of ≥3 diseased coronary arteries. In contrast, individual vulnerable plaque features, including LRNC, IPH, LM, and ulceration, showed weaker and less consistent associations. Conclusions: In patients undergoing carotid revascularization, preoperative carotid HR-MRI-derived plaque burden-related features showed more consistent associations with coronary atherosclerotic burden than individual vulnerable plaque features. These findings suggest that plaque burden-related and vulnerable plaque features may reflect different aspects of atherosclerosis, with plaque burden-related measures more closely reflecting cumulative systemic atherosclerotic involvement and vulnerable plaque features primarily characterizing local plaque composition and instability.

Background/Objectives: Patients undergoing carotid revascularization may show distinct patterns of bilateral carotid vulnerability and clinical risk. This study used latent class analysis (LCA) to identify high-resolution magnetic resonance imaging (HR-MRI)-based clinical–radiological phenotypes and evaluate their association with long-term major adverse cardiovascular events (MACE). Methods: This retrospective cohort included 289 patients who underwent carotid endarterectomy (CEA) or carotid artery stenting (CAS) between April 2017 and April 2024 and had preoperative carotid HR-MRI within 90 days. Ipsilateral and contralateral plaque features were assessed on multi-contrast vessel-wall HR-MRI. LCA included five indicators: ipsilateral intraplaque hemorrhage, ipsilateral lipid-rich necrotic core, contralateral intraplaque hemorrhage, cardiovascular disease history, and symptomatic status. MACE comprised cardiovascular death, nonfatal myocardial infarction, coronary revascularization, or stroke. Outcomes were assessed using Kaplan–Meier analysis and multivariable Cox regression. Results: Among 289 patients undergoing carotid revascularization, the mean age was 65.12 ± 9.46 years, and 237 (82.0%) patients were men. LCA identified three phenotypes: Systemic High-Risk (n = 128), Silent High-Risk (n = 37), and Stable/Moderate (n = 124). The Systemic High-Risk phenotype showed high probabilities of symptomatic presentation, cardiovascular disease history, and bilateral carotid vulnerability, whereas the Silent High-Risk phenotype showed marked ipsilateral plaque vulnerability despite lower clinical risk. During a median follow-up of 4.4 years, MACE-free survival differed significantly across phenotypes (log-rank p = 0.043). Compared with the Stable/Moderate phenotype, the Systemic High-Risk and Silent High-Risk phenotypes were both associated with increased MACE risk (HR 1.65, 95% CI 1.15–2.37, p = 0.039; HR 1.43, 95% CI 1.12–2.15, p = 0.034). Conclusions: HR-MRI-based LCA identified clinically meaningful carotid vulnerability phenotypes associated with long-term MACE after carotid revascularization. These findings support integrated assessment of bilateral plaque vulnerability and clinical risk for postoperative cardiovascular risk stratification.

Purpose: To evaluate the prognostic value of carotid plaque features derived from high-spatial-resolution vessel-wall MRI (HR-MRI) for long-term major adverse cardiovascular events (MACE) after carotid revascularization. Methods: Consecutive patients undergoing carotid revascularization between April 2017 and April 2024 with preoperative carotid HR-MRI were included. Ipsilateral intraplaque hemorrhage (IPH) was identified as a hyperintense plaque component on SNAP images and manually segmented on each relevant slice using Vessel Explorer 2.0 software. IPH volume was calculated as the sum of the segmented IPH areas multiplied by slice thickness. MACE comprised cardiovascular death, nonfatal myocardial infarction, coronary revascularization, and stroke. Associations were assessed using Cox regression and Kaplan–Meier analysis. Prediction models were evaluated using the concordance index (C-index), calibration, and decision curve analysis (DCA). Results: Among 296 patients (mean age [65.04 ± 9.47] years; 244 [82.4%] men), 154 underwent carotid endarterectomy (CEA) and 142 underwent carotid artery stenting (CAS). During a median follow-up of 4.5 years, ipsilateral IPH volume was an independent predictor of MACE in the overall cohort (hazard ratio [HR], 1.49; 95% confidence interval [CI], 1.34-1.66; P < .001), as well as in the CEA subgroup (HR, 1.31; 95% CI, 1.11-1.55; P = .001) and CAS subgroup (HR, 1.21; 95% CI, 1.05-1.40; P = .007). Contralateral IPH was also independently associated with higher event risk in the overall cohort (HR, 2.37; 95% CI, 1.34-4.19; P = .003) and in both procedural subgroups (CEA: HR, 2.79; 95% CI, 1.16-6.67; P = .022; CAS: HR, 2.47; 95% CI, 1.22-5.00; P = .012). Kaplan–Meier analysis showed significantly lower event-free survival in patients with ipsilateral or contralateral IPH. Prediction models demonstrated acceptable discrimination (C-index: overall, 0.725; CEA, 0.755; CAS, 0.711), good calibration, and consistent clinical net benefit. Conclusions: Among HR-MRI–derived carotid plaque features, ipsilateral IPH volume and the presence of contralateral IPH were independently associated with long-term cardiovascular events after carotid revascularization. These findings support carotid IPH as an imaging marker of systemic atherosclerotic vulnerability and may improve individualized risk stratification after revascularization.

Background: Artificial intelligence (AI), particularly deep learning, has shown promise in enhancing medical image interpretation and improving radiologists’ efficiency. In China, growing imaging demand and workforce shortages have placed increasing pressure on radiology services. However, evidence on the operational impact of AI on reporting efficiency remains limited. Objective: This study aimed to evaluate the effect of an AI system on radiologists’ reporting efficiency by examining changes in report-drafting time for lung nodule diagnosis in chest computed tomography (CT) images. Methods: We analyzed 185,044 chest CT reports from Beijing Anzhen and Tsinghua Changgung Hospitals (2018-2023) using a difference-in-differences design with nonequivalent comparison groups. Report-drafting time before, immediately after, and up to 2 years following AI implementation was compared, adjusting for radiologist gender, seniority, and years of working experience. Results: The pooled analysis showed a modest overall increase of 0.86 minutes (95% CI 0.14 to 1.57). However, this masked substantial heterogeneity between hospitals due to differing implementation timelines. In the first year after AI deployment, Tsinghua Changgung Hospital experienced a nonsignificant increase of 0.90 minutes (95% CI –0.28 to 2.08). In contrast, at Beijing Anzhen Hospital, the AI-assisted group exhibited an absolute reduction of 0.76 minutes by the first year and a further 1.83-minute reduction by the second year (an approximate 28% time saved vs baseline), while the control group remained stable over time. Using a difference-in-differences framework, this corresponded to a 2.66-minute relative improvement compared with the counterfactual trend (P<.001). Conclusions: AI-assisted lung nodule diagnosis may initially increase report-drafting time due to adaptation and workflow adjustment. Sustained, meaningful efficiency gains were heterogeneous and observed at only 1 of the 2 study sites, indicating that long-term impacts are strongly contingent on site-specific implementation dynamics, learning curves, and local context.

Objectives: It is well established that calcified plaques are highly likely to lead to residual stenosis after stenting; however, the specific characteristics responsible for this effect remain unknown. This study aimed to identify both qualitative and quantitative imaging risk factors for residual stenosis using computed tomography angiography. Methods: We retrospectively enrolled 233 patients with carotid artery stenosis. Patients were categorized into two groups based on the presence or absence of postoperative residual stenosis. Carotid computed tomography angiography evaluated plaque characteristics both qualitatively and quantitatively. Logistic regression analysis identified independent risk factors for residual stenosis. We evaluated the predictive model’s discriminative ability by calculating the area under the receiver operating characteristic (ROC) curve. Results: Univariate analysis indicated a statistical difference in age, creatinine, total plaque volume, percentage of total calcified plaque, percentage of total soft plaque, maximum slice attenuation value, maximum thickness, total length, and a circumferential calcification score ≥2 points ( p < 0.05). Multivariable logistic regression identified creatinine (OR = 1. 020; 95%CI: 1.005–1.035; p = 0.010), maximum slice attenuation value(Z-score; OR = 1.627; 95%CI: 1.024–2.585; p = 0.039), percentage of calcified plaque volume(Z-score; OR = 1.872; 95%CI: 1.137–3.082; p = 0.014) and circumferential calcification score ≥2 (OR = 3.257; 95%CI: 1.620–6.548; p < 0.001) as independent factors associated with residual stenosis. Furthermore, receiver operating characteristic curve analysis revealed that the area under the curve for the combined model in diagnosing residual stenosis was 0.784. Conclusion: In conclusion, preoperative CTA-based assessment of specific plaque characteristics, such as calcified plaque volume percentage, circumferential calcium score, and the maximum slice attenuation value of calcification are related to residual stenosis.

BACKGROUND Artificial intelligence (AI), particularly deep learning, has shown promise in enhancing medical image interpretation and improving radiologists’ efficiency. In China, growing imaging demand and workforce shortages have placed increasing pressure on radiology services. However, evidence on the operational impact of AI on reporting efficiency remains limited. OBJECTIVE This study aimed to evaluate the effect of an AI system on radiologists’ reporting efficiency by examining changes in report-drafting time for lung nodule diagnosis in chest computed tomography (CT) images. METHODS We analyzed 185,044 chest CT reports from Beijing Anzhen and Tsinghua Changgung Hospitals (2018-2023) using a difference-in-differences design with nonequivalent comparison groups. Report-drafting time before, immediately after, and up to 2 years following AI implementation was compared, adjusting for radiologist gender, seniority, and years of working experience. RESULTS The pooled analysis showed a modest overall increase of 0.86 minutes (95% CI 0.14 to 1.57). However, this masked substantial heterogeneity between hospitals due to differing implementation timelines. In the first year after AI deployment, Tsinghua Changgung Hospital experienced a nonsignificant increase of 0.90 minutes (95% CI –0.28 to 2.08). In contrast, at Beijing Anzhen Hospital, the AI-assisted group exhibited an absolute reduction of 0.76 minutes by the first year and a further 1.83-minute reduction by the second year (an approximate 28% time saved vs baseline), while the control group remained stable over time. Using a difference-in-differences framework, this corresponded to a 2.66-minute relative improvement compared with the counterfactual trend (P<.001). CONCLUSIONS AI-assisted lung nodule diagnosis may initially increase report-drafting time due to adaptation and workflow adjustment. Sustained, meaningful efficiency gains were heterogeneous and observed at only 1 of the 2 study sites, indicating that long-term impacts are strongly contingent on site-specific implementation dynamics, learning curves, and local context.

Background: Uncertainty in the diagnosis of lung nodules is a challenge for both patients and physicians. Artificial intelligence (AI) systems are increasingly being integrated into medical imaging to assist diagnostic procedures. However, the accuracy of AI systems in identifying and measuring lung nodules on chest computed tomography (CT) scans remains unclear, which requires further evaluation. Objective: This study aimed to evaluate the impact of an AI-assisted diagnostic system on the diagnostic efficiency of radiologists. It specifically examined the report modification rates and missed and misdiagnosed rates of junior radiologists with and without AI assistance. Methods: We obtained effective data from 12,889 patients in 2 tertiary hospitals in Beijing before and after the implementation of the AI system, covering the period from April 2018 to March 2022. Diagnostic reports written by both junior and senior radiologists were included in each case. Using reports by senior radiologists as a reference, we compared the modification rates of reports written by junior radiologists with and without AI assistance. We further evaluated alterations in lung nodule detection capability over 3 years after the integration of the AI system. Evaluation metrics of this study include lung nodule detection rate, accuracy, false negative rate, false positive rate, and positive predictive value. The statistical analyses included descriptive statistics and chi-square, Cochran-Armitage, and Mann-Kendall tests. Results: The AI system was implemented in Beijing Anzhen Hospital (Hospital A) in January 2019 and Tsinghua Changgung Hospital (Hospital C) in June 2021. The modification rate of diagnostic reports in the detection of lung nodules increased from 4.73% to 7.23% (χ21=12.15; P<.001) at Hospital A. In terms of lung nodule detection rates postimplementation, Hospital C increased from 46.19% to 53.45% (χ21=25.48; P<.001) and Hospital A increased from 39.29% to 55.22% (χ21=122.55; P<.001). At Hospital A, the false negative rate decreased from 8.4% to 5.16% (χ21=9.85; P=.002), while the false positive rate increased from 2.36% to 9.77% (χ21=53.48; P<.001). The detection accuracy demonstrated a decrease from 93.33% to 92.23% for Hospital A and from 95.27% to 92.77% for Hospital C. Regarding the changes in lung nodule detection capability over a 3-year period following the integration of the AI system, the detection rates for lung nodules exhibited a modest increase from 54.6% to 55.84%, while the overall accuracy demonstrated a slight improvement from 92.79% to 93.92%. Conclusions: The AI system enhanced lung nodule detection, offering the possibility of earlier disease identification and timely intervention. Nevertheless, the initial reduction in accuracy underscores the need for standardized diagnostic criteria and comprehensive training for radiologists to maximize the effectiveness of AI-enabled diagnostic systems.