Automatic identification of coronary stent in coronary calcium scoring CT using deep learning

Coronary artery calcium (CAC) scoring CT is a useful tool for screening coronary artery disease and for cardiovascular risk stratification. However, its efficacy in patients with coronary stents, who had pre-existing coronary artery disease, remains uncertain. Historically, CAC CT scans of these patients have been manually excluded from the CAC scoring process, even though most of the CAC scoring process is now fully automated. Therefore, we hypothesized that automating the filtering of patients with coronary stents using artificial intelligence could streamline the entire CAC workflow, eliminating the need for manual intervention. Consequently, we aimed to develop and evaluate a deep learning-based coronary stent filtering algorithm (StentFilter) in CAC scoring CT scans using a multicenter CAC dataset. We developed StentFilter comprising two main processes: stent identification and false-positive reduction. Development utilized 108 non-enhanced echocardiography-gated CAC scans (including 74 with manually labeled stents), and for false positive reduction, 2063 CAC scans with significant coronary calcium (average Agatston score: 523.8) but no stents were utilized. StentFilter's performance was evaluated on two independent internal test sets (Asan cohort- and 2; n = 355 and 396; one without coronary stents) and two external test sets from different institutions (n = 105 and 62), each with manually labeled stents. We calculated the per-patient sensitivity, specificity, and false-positive rate of StentFilter. StentFilter demonstrated a high overall per-patient sensitivity of 98.8% (511/517 cases with stents) and a false-positive rate of 0.022 (20/918). Notably, the false-positive ratio was significantly lower in the dataset containing stents (Asan cohort-1; 0.008 [3/355]) compared with the dataset without stents (Asan cohort-2; 0.043 [17/396], p = 0.008). All false-positive identifications were attributed to dense coronary calcifications, with no false positives identified in extracoronary locations. The automated StentFilter accurately distinguished coronary stents from pre-existing coronary calcifications. This approach holds potential as a filter within a fully automated CAC scoring workflow, streamlining the process efficiently.

Biomarkers for identification of high-risk coronary artery plaques in patients with suspected coronary artery disease

BACKGROUND

Patients with atherosclerotic plaques containing high-risk features have an increased likelihood of events and a worse prognosis. Whether increased levels of Troponin I (TnI) and C-reactive protein (CRP) are associated with the presence of high-risk coronary atherosclerotic plaques (HRP) is not well described. We assessed the association between 1) TnI and 2) CRP with quantified coronary plaque burden, luminal diameter stenosis, and HRP in patients with low/intermediate pre-test probability of obstructive coronary artery disease (CAD) referred for coronary computed tomography angiography (CCTA).

METHODS

The CCTA from 1615 patients were analyzed using a semiautomatic software for coronary artery plaque characterization. Patients with high TnI (>6 ​ng/L) and high CRP (>2 ​mg/L) were identified. Associations of TnI and CRP with plaque burden, stenosis (≥50% luminal diameter stenosis on CCTA), and HRP were investigated.

RESULTS

TnI and CRP were both positively correlated with total plaque burden (TnI rs ​= ​0.14, p ​< ​0.001; CRP rs ​= ​0.08, p ​< ​0.001). In multivariate logistic regression analyses, high TnI was associated with stenosis (OR 1.43, 95% confidence interval (CI) 1.03-1.99, p ​= ​0.034), the presence of HRP (OR 1.79, 95% CI: 1.17-2.74, p ​= ​0.008), and the subtypes of HRP; low attenuation plaque (OR 1.93, 95% CI: 1.24-3.00, p ​= ​0.003), and positive remodeling (OR 1.51, 95% CI: 1.07-2.13, p ​= ​0.018). For CRP, only stenosis and napkin ring sign correlated significantly.

CONCLUSION

In patients with suspected CAD, TnI and CRP are associated with HRP features. These findings may suggest that inflammatory and particularly ischemic biomarkers might improve early risk stratification and affect patient management.

CLINICALTRIALS

GOV IDENTIFIER

NCT02264717.

Assessment of Systemic Immune-Inflammation Index and Systemic Inflammation-Response Index in Different Coronary Artery Plaque Types

Coronary computed tomography angiography can evaluate coronary arterial plaque composition with high resolution. We aimed to determine and compare the values of systemic immune-inflammation index (SII) and systemic inflammation-response index (SIRI) in different plaque types. Highest values of SIRI and SII were found in mixed plaque type followed by non-calcified plaque type. SII value of 463.07 predicted one-year major adverse cardiac events (MACE) with a sensitivity of 72.7% and specificity of 64.3% and SIRI value of 1.14 predicted one-year MACE with a sensitivity of 93% and specificity of 62%. Paired analysis of area under the curve (AUC) of the receiver operating characteristic curves showed that SIRI had higher AUC than coronary calcium score and SII. Univariate logistic regression results showed that age, creatinine level, coronary calcium score, SII, and SIRI were the independent predictors of one-year MACE. According to the results of multivariate regression analysis, after adjusting other factors, age, creatinine level, and SIRI were the independent predictors of one-year MACE. SIRI seemed to improve the risk prediction in coronary artery disease. Therefore, special attention may need to be paid to patients who have a high SIRI.

Fully Automatic Coronary Calcium Score Software Empowered by Artificial Intelligence Technology: Validation Study Using Three CT Cohorts

Objective

This study aimed to validate a deep learning-based fully automatic calcium scoring (coronary artery calcium [CAC]_auto) system using previously published cardiac computed tomography (CT) cohort data with the manually segmented coronary calcium scoring (CAC_hand) system as the reference standard.

Materials and Methods

We developed the CAC_auto system using 100 co-registered, non-enhanced and contrast-enhanced CT scans. For the validation of the CAC_auto system, three previously published CT cohorts (n = 2985) were chosen to represent different clinical scenarios (i.e., 2647 asymptomatic, 220 symptomatic, 118 valve disease) and four CT models. The performance of the CAC_auto system in detecting coronary calcium was determined. The reliability of the system in measuring the Agatston score as compared with CAC_hand was also evaluated per vessel and per patient using intraclass correlation coefficients (ICCs) and Bland-Altman analysis. The agreement between CAC_auto and CAC_hand based on the cardiovascular risk stratification categories (Agatston score: 0, 1–10, 11–100, 101–400, > 400) was evaluated.

Results

In 2985 patients, 6218 coronary calcium lesions were identified using CAC_hand. The per-lesion sensitivity and false-positive rate of the CAC_auto system in detecting coronary calcium were 93.3% (5800 of 6218) and 0.11 false-positive lesions per patient, respectively. The CAC_auto system, in measuring the Agatston score, yielded ICCs of 0.99 for all the vessels (left main 0.91, left anterior descending 0.99, left circumflex 0.96, right coronary 0.99). The limits of agreement between CAC_auto and CAC_hand were 1.6 ± 52.2. The linearly weighted kappa value for the Agatston score categorization was 0.94. The main causes of false-positive results were image noise (29.1%, 97/333 lesions), aortic wall calcification (25.5%, 85/333 lesions), and pericardial calcification (24.3%, 81/333 lesions).

Conclusion

The atlas-based CAC_auto empowered by deep learning provided accurate calcium score measurement as compared with manual method and risk category classification, which could potentially streamline CAC imaging workflows.

C-reactive protein-to-serum albumin ratio as a novel predictor of long-term outcomes in coronary artery disease patients who have undergone percutaneous coronary intervention: analysis of a real-world retrospective cohort study

BACKGROUND

C-reactive protein (CRP) has been proposed as a contributor to the pathogenesis of coronary artery disease (CAD) and inflammatory reactions, which are associated with a decrease in serum albumin, and it has been reported that the CRP-to-serum albumin ratio (CAR) can predict CAD severity in inpatient ischemic cardiomyopathy (ICM) patients. However, the relationship between the CAR and long-term adverse outcomes in CAD patients after percutaneous coronary intervention (PCI) is still unknown.

METHODS

A total of 3561 CAD patients enrolled in the Outcomes and Risk Factors of Patients with Coronary Heart Disease after PCI: an investigation based on case records and follow-up (CORFCHD-ZZ), a retrospective cohort study conducted from January 2013 to December 2017, and 1630 patients meeting the study inclusion criteria were divided into two groups based on the CAR (CAR < 0.186; n = 1301 and CAR ≥ 0.186; n = 329). The primary outcome was long-term mortality, including all-cause mortality (ACM) and cardiac mortality. The average follow-up time was 37.59 months.

RESULTS

We found that there were significant differences between the two groups in the incidences of ACM (P < 0.001) and cardiac mortality (P  = 0.003). Cox multivariate regression analyses demonstrated that CAR was an independent predictor of ACM [hazard ratio, 2.678; (95% confidence interval (CI), 1.568-4.576); P < 0.001] and cardiac mortality (hazard ratio, 2.055; 95% CI, 1.056-3.998; P = 0.034) in CAD patients after PCI.

CONCLUSION

This study revealed that the CAR is an independent and novel predictor of long-term adverse outcomes in CAD patients who have undergone PCI.

The correlation between hepatic fat fraction evaluated by dual-energy computed tomography and high-risk coronary plaques in patients with non-alcoholic fatty liver disease

PURPOSE

To determine the relationship between non-alcoholic fatty liver disease (NAFLD) evaluated by a hepatic fat fraction (HFF) using dual-energy computed tomography (DECT) and high-risk coronary plaques (HRP) in NAFLD patients.

METHODS

We conducted a matched case-control study involving 172 NAFLD individuals recruited from August 2019 to September 2020. They underwent dual-energy coronary computed tomographic angiography and were classified as no-plaque, HRP negative and HRP positive groups. HFF values were measured using multimaterial decomposition algorithm of DECT, and the differences among three groups were compared. Multiple logistic regression analysis was performed to determine the independent correlation between HFF and HRP. Spearman rank correlation was used to assess the correlations between HFF and multiple variables.

RESULTS

HRP positive group (15.3%) had higher HFF values than no-plaque (6.9%) and HRP negative groups (8.9%) (P < 0.001). After adjusting for confounding variables, the results indicated that HFF was an independent risk factor for HRP (OR 1.93, P < 0.001). Additionally, HFF significantly correlated with coronary artery calcium score, hepatic CT attenuation, epicardial and pericoronary adipose tissue volume, and CT attenuation (all P < 0.001).

CONCLUSIONS

As a new imaging marker for the quantification of liver fat, HFF was independently associated with HRP.

Practice Pattern, Diagnostic Yield, and Long-Term Prognostic Impact of Coronary Computed Tomographic Angiography

Background Although guidelines recommend the use of coronary computed tomographic angiography (CTA) in patients with stable pain syndromes, the clinical benefits of the use of coronary CTA in a broad spectrum of patients is unknown. We evaluated the contemporary practice pattern and diagnostic yield of coronary CTA and their impact on the subsequent diagnostic-therapeutic cascade and clinical outcomes. Methods and Results We identified 39 906 patients without known coronary artery disease (CAD) who underwent coronary CTA between January 2007 and December 2013. The patients' demographic characteristics, risk factors, symptoms, results of coronary CTA, the appropriateness of downstream diagnostic and therapeutic interventions, and long-term outcomes (death or myocardial infarction) were evaluated. The number of coronary CTAs had increased over time, especially in asymptomatic patients. Coronary CTA revealed that 6108 patients (15.3%) had obstructive CAD (23.7% of symptomatic and 9.3% of asymptomatic patients). Subsequent cardiac catheterization was performed in 19.2% of symptomatic patients (appropriate, 80.6%) and in 3.9% of asymptomatic patients (appropriate, 7.9%). The 5-year rate of death or myocardial infarction was significantly higher in patients with obstructive CAD on CTA than those without (7.2% versus 3.0%; P<0.001; adjusted hazard ratio [95% CI], 1.34 [1.17-1.54]). However, obstructive CAD on CTA had limited added value over conventional risk factors for predicting death or myocardial infarction. Conclusions Although the use of coronary CTA had substantially increased, CTA had a low diagnostic yield for obstructive CAD, especially in asymptomatic patients. The use of CTA in asymptomatic patients seemed to have led to inappropriate subsequent diagnostic or therapeutic interventions without clinical benefit.

Age-Related Differences in the Association between Plasma High-Sensitivity C-Reactive Protein and Noncalcified or Mixed Coronary Atherosclerotic Plaques

Background

Previous studies have demonstrated that plasma high-sensitivity C-reactive protein (hsCRP) was the predictor for unstable coronary plaque. Patients with noncalcified plaque (NCP) or mixed plaque (MP) have a higher risk of poor outcomes. However, the association between hsCRP and the presence of NCP or MP (NCP/MP) in old adults remains unclear, and if present, whether there exist differences between young and old adults remain unknown. Thus, the aim of this study was to investigate the role of hsCRP in predicting the presence of NCP/MP and evaluate whether age has any impact on this association.

Methods

A total of 951 subjects were included in this study. Complete clinical and laboratory data were collected. According to the characteristics of the most stenotic plaque, we divided them into 2 groups: calcified plaque (CP) and NCP/MP. Subjects with no plaque were classified as the control group (CR). Subjects with age ≥ 60 years were defined as older adults, and those with age < 60 years were classified as nonelderly people.

Results

Patients with NCP/MP had significantly higher hsCRP level compared with subjects with CR or CP in older adults but not in nonelderly people. The proportion of NCP/MP was significantly increased from 27.0% in the hsCRP < 1.25 mg/L group to 42.7% in the hsCRP > 2.70 mg/L group in older adults. Multiple logistic regression analysis showed that hsCRP was an independent risk factor for the presence of NCP/MP (odds ratio (OR) = 1.093, 95% CI 1.032–1.157, P = 0.001) only in older adults.

Conclusions

hsCRP is independently associated with the presence of NCP/MP in older adults but not in nonelderly people. These results suggest the potential significance of hsCRP-lowering regimens in older adults with NCP/MP.

Personalized Statin Therapy and Coronary Atherosclerotic Plaque Burden in Asymptomatic Low/Intermediate-Risk Individuals

BACKGROUND

Current guidelines for the primary prevention of atherosclerotic cardiovascular disease are based on the estimation of a predicted 10-year cardiovascular disease risk and the average relative risk reduction estimates from statin trials. In the clinical setting, however, decision-making is better informed by the expected benefit for the individual patient, which is typically lacking. Consequently, a personalized statin benefit approach based on absolute risk reduction over 10 years (ARR10 benefit threshold ≥2.3%) has been proposed as a novel approach. However, how this benefit threshold relates with coronary plaque burden in asymptomatic individuals with low/intermediate cardiovascular disease risk is unknown.

AIMS

In this study, we compared the predicted ARR10 obtained in each individual with plaque burden detected by coronary computed tomography angiography.

METHODS AND RESULTS

Plaque burden (segment volume score, segment stenosis score, and segment involvement score) was assessed in prospectively recruited asymptomatic subjects (n = 70; 52% male; median age 56 years [interquartile range 51-64 years]) with low/intermediate Framingham risk score (< 20%). The expected ARR10 with statin in the entire cohort was 2.7% (1.5-4.6%) with a corresponding number needed to treat over 10 years of 36 (22-63). In subjects with an ARR10 benefit threshold ≥2.3% (vs. < 2.3%), plaque burden was significantly higher (p = 0.02).

CONCLUSION

These findings suggest that individuals with higher coronary plaque burden are more likely to get greater benefit from statin therapy even among asymptomatic individuals with low cardiovascular risk.

Association between high sensitivity C-Reactive protein and prevalence of asymptomatic carotid artery stenosis

INTRODUCTION

Inflammation plays a key role in the pathophysiology of atherosclerosis. Little is known about the association between high sensitivity C-reactive protein (hs-CRP) especially long-term hs-CRP and asymptomatic carotid artery stenosis (ACAS) in healthy Chinese adults. The aim of this study was to assess the relationship between hs-CRP levels and the prevalence of ACAS in a Chinese community-based cohort.

METHODS

A sample of 5349 participants aged ≥ 40 years (40.36% women) were enrolled in this study, all without preexisting stroke. Ultrasonography of the bilateral carotid arteries was performed for the evaluation of carotid stenosis. Participants were stratified into three groups according to hs-CRP levels. We used both baseline (hs-CRP levels analyzed during 2010) and average hs-CRP values for the last four years (the average of hs-CRP levels analyzed at the year of 2006, 2008 and 2010) in the analysis. Multivariable logistic regression models were used to analyze the association between hs-CRP levels and ACAS.

RESULTS

A total of 356 (6.66%) subjects showed evidence of ACAS. Multivariate analysis showed that both baseline and average hs-CRP values for the last four years were independent indicators for the presence of ACAS (P for trend = 0.007, 0.001, respectively). Stratified by age and sex, higher baseline hs-CRP levels were associated with ACAS in old adults (≥ 60 y) (multivariate-adjusted, odds ratio [OR] = 1.03, 95% confidence interval [CI]: 1.01-1.05) and male (multivariate-adjusted, OR = 1.03, 95%CI: 1.01-1.05), but not in middle-aged adults (40-59 y) and female. Similarly, higher average hs-CRP values for the last four years were associated with ACAS in old adults and male, but not in middle-aged adults and female.

CONCLUSION

Both baseline and chronic elevation of serum hs-CRP were associated with ACAS, especially in older or male adults. hs-CRP might be used as a useful marker and a potential therapeutic target for carotid atherosclerosis.

Coronary Plaque Burden at Coronary CT Angiography in Asymptomatic Men and Women

Purpose To assess the relationship between total, calcified, and noncalcified coronary plaque burdens throughout the entire coronary vasculature at coronary computed tomographic (CT) angiography in relationship to cardiovascular risk factors in asymptomatic individuals with low-to-moderate risk. Materials and Methods This HIPAA-compliant study had institutional review board approval, and written informed consent was obtained. Two hundred two subjects were recruited to an ongoing prospective study designed to evaluate the effect of HMG-CoA reductase inhibitors on atherosclerosis. Eligible subjects were asymptomatic individuals older than 55 years who were eligible for statin therapy. Coronary CT angiography was performed by using a 320-detector row scanner. Coronary wall thickness and plaque were evaluated in all epicardial coronary arteries greater than 2 mm in diameter. Images were analyzed by using dedicated software involving an adaptive lumen attenuation algorithm. Total plaque index (calcified plus noncalcified plaque) was defined as plaque volume divided by vessel length. Multivariable regression analysis was performed to determine the relationship between risk factors and plaque indexes. Results The mean age of the subjects was 65.5 years ± 6.9 (standard deviation) (36% women), and the median coronary artery calcium (CAC) score was 73 (interquartile range, 1-434). The total coronary plaque index was higher in men than in women (42.06 mm(2) ± 9.22 vs 34.33 mm(2) ± 8.35; P < .001). In multivariable analysis controlling for all risk factors, total plaque index remained higher in men than in women (by 5.01 mm(2); P = .03) and in those with higher simvastatin doses (by 0.44 mm(2)/10 mg simvastatin dose equivalent; P = .02). Noncalcified plaque index was positively correlated with systolic blood pressure (β = 0.80 mm(2)/10 mm Hg; P = .03), diabetes (β = 4.47 mm(2); P = .03), and low-density lipoprotein (LDL) cholesterol level (β = 0.04 mm(2)/mg/dL; P = .02); the association with LDL cholesterol level remained significant (P = .02) after additional adjustment for the CAC score. Conclusion LDL cholesterol level, systolic blood pressure, and diabetes were associated with noncalcified plaque burden at coronary CT angiography in asymptomatic individuals with low-to-moderate risk. (©) RSNA, 2015 Online supplemental material is available for this article.