Cardiac Cardiac Computed Tomography Research Articles

Deep learning-based reconstruction on cardiac CT yields distinct radiomic features compared to iterative and filtered back projection reconstructions

We aimed to determine the effects of deep learning-based reconstruction (DLR) on radiomic features obtained from cardiac computed tomography (CT) by comparing with iterative reconstruction (IR), and filtered back projection (FBP). A total of 284 consecutive patients with 285 cardiac CT scans that were reconstructed with DLR, IR, and FBP, were retrospectively enrolled. Radiomic features were extracted from the left ventricular (LV) myocardium, and from the periprosthetic mass if patients had cardiac valve replacement. Radiomic features of LV myocardium from each reconstruction were compared using a fitting linear mixed model. Radiomics models were developed to diagnose periprosthetic abnormality, and the performance was evaluated using the area under the receiver characteristics curve (AUC). Most radiomic features of LV myocardium (73 of 88) were significantly different in pairwise comparisons between all three reconstruction methods (P < 0.05). The radiomics model on IR exhibited the best diagnostic performance (AUC 0.948, 95% CI 0.880–1), relative to DLR (AUC 0.873, 95% CI 0.735–1) and FBP (AUC 0.875, 95% CI 0.731–1), but these differences did not reach significance (P > 0.05). In conclusion, applying DLR to cardiac CT scans yields radiomic features distinct from those obtained with IR and FBP, implying that feature robustness is not guaranteed when applying DLR.

Contrast defect of left atrial appendage on computed tomography is associated with higher risk of recurrence after catheter ablation in patients with paroxysmal atrial fibrillation.

The previous study has shown that the contrast defect of the left atrial appendage (LAA) on contrast-enhanced cardiac computed tomography (CT) is associated with a higher rate of stroke in patients with atrial fibrillation (AF). This study aimed to investigate the association between LAA CT contrast defect and the risk of arrhythmia recurrence after catheter ablation (CA) in patients with paroxysmal AF. A total of 283 paroxysmal AF patients [age: 67 ± 10years, 185 (65%) males] who underwent cardiac CT before CA were retrospectively analyzed. The presence or absence of LAA CT contrast defect was visually assessed using early phase CT images. Recurrence was an episode of atrial arrhythmia beyond the first 90days post-ablation. LAA flow velocity was measured using transesophageal echocardiography in 246 paroxysmal AF patients. Sixty-eight (24%) patients had an LAA CT contrast defect. LAA flow velocity was significantly reduced in patients with LAA CT defect compared to those without (56.8 ± 28.7cm/s vs. 41.1 ± 19.1cm/s, p < 0.001). During a median follow-up period of 858days, arrhythmia recurrence was identified in 85 (30%) patients. On a Kaplan Meier curve, patients with LAA CT contrast defect had significantly higher recurrence rates than those without (p = 0.043). On a multivariable Cox regression analysis, LAA CT contrast defect was a significant and independent predictor after adjustment of age, sex and left atrial volume index (hazard ratio: 1.79, 95% confidence interval: 1.03-3.07, p = 0.036). LAA CT contrast defect was associated with decreased LAA flow velocity and a higher rate of arrhythmia recurrence after CA, suggesting its usefulness as a non-invasive predictor for high-risk AF patients resistant to CA therapy.

Is the Calcium Score Useful for Rheumatoid Arthritis Patients at Low or Intermediate Cardiovascular Risk?

Cardiovascular disease, particularly myocardial infarction, is the leading cause of death of rheumatoid arthritis (RA) patients. The usefulness of the coronary artery calcification score (CACS), determined using cardiac computed-tomography (CT)-scan images, was assessed as a part of a cardiovascular work-up of RA patients at low or intermediate cardiovascular disease risk. This descriptive, cross-sectional, single-center study was conducted on patients with stable RA or that which is in remission. Each patient’s work-up included a collection of cardiovascular risk factors, laboratory analyses, an electrocardiogram, a supra-aortic trunks (SATs) echo-Doppler test and a cardiac CT scan. The primary endpoint was to determine the frequency of patients with a CACS > 100, indicating notable atherosclerosis. Fifty patients were analyzed: mean ± standard deviation age was 53.7 ± 7.5 years, 82% women. The CACS exceeded 100 in 12 (24%) patients (11 were at intermediate risk) and 2 of them underwent angioplasty for silent myocardial ischemia. Cardiovascular risk was reclassified from intermediate to high for 5 patients. Age according to sex and smoking status were significantly associated with that increase; no association was found with RA characteristics or treatments.

Evaluation of extracellular volume by computed tomography is useful for prediction of prognosis in dilated cardiomyopathy.

Cardiac computed tomography (CT) is useful for the screening of coronary artery stenosis, and extracellular volume fraction (ECV) analysis by CT using new dedicated software is now available. Here, we evaluated the utility of ECV analysis using cardiac CT to predict patient prognosis in cases with dilated cardiomyopathy (DCM). We analyzed 70 cases with DCM and cardiac computed tomography (CT) with available late-phase images. We evaluated the ECV of the left ventricular myocardium (LVM) using commercially available software (Ziostation 2, Ziosoft Inc, Japan). ECV on LVM was 33.96 ± 5.04%. Major adverse cardiac events (MACE) occurred in 21 cases (30%). ECV of the LVM on CT, endo-systolic volume, and rate of significant valvular disease were significantly higher in cases with MACE than in those without (37.16 ± 5.91% vs. 32.59 ± 3.95%, 194 ± 109 vs. 138 ± 78ml and 57% vs. 20%, all P values < 0.05). LVEF was significantly lower in cases with MACE than in those without (23 ± 8 vs. 31 ± 11%, P = 0.0024). The best cut-off value of ECV on LVM for prediction of MACE was 32.26% based on receiver operating characteristics analysis. Cases with ECV ≥ 32.26% had significantly higher MACE based on Kaplan-Meier analysis (P = 0.0032). Only ECV on LVM was an independent predictor of MACE based on a multivariate Cox proportional hazards model (P = 0.0354). Evaluation of ECV on LVM by CT is useful for predicting MACE in patients with DCM.

Left atrial appendage sealing performance of the Amplatzer Amulet and Watchman FLX device

BackgroundThe left atrial appendage (LAA) sealing properties of the Amplatzer Amulet and Watchman FLX devices were compared using cardiac computed tomography (CT) follow-up.MethodsSingle-center cohort study of patients undergoing LAAO between 2017 and 2020. Two consecutive cohorts were enrolled, one treated with the Amplatzer Amulet (n = 150) up till 2019, and a second cohort treated with the Watchman FLX (n = 150) device from 2019. Cardiac CT was performed 2 months postprocedure. The primary outcome was complete LAA occlusion defined as no visible peri-device leak (PDL) and absence of contrast patency in the distal LAA. Secondary outcomes included PDL, contrast patency without visible PDL, PDL area, and periprocedural complications.ResultsComplete occlusion was achieved in 39 (30.5%) of the Amulet group, compared to 89 (71.8%) of the FLX group, p < 0.001. A PDL at the Amulet disc was present in 65 (50.8%), at the lobe in 16 (12.5%), and at both the disc and lobe in 13 (10.2%). For FLX, a PDL was present in 20 (16.1%). Contrast patency without visible PDL was observed in 24 (18.8%) and 15 (12.1%) of the Amulet and FLX group, respectively. The PDL area at the Amulet mid-lobe was 92 mm2 (59–158) and 32 mm2 (IQR 28–96) for FLX, p = 0.019. Device-related thrombosis occurred in 1 (0.7%) and 2 (1.3%), respectively (p = 0.99), with periprocedural adverse events occurring in 6 (4%) and 8 (5.3%) of the Amulet and FLX group (p = 0.79).ConclusionComplete LAA occlusion was achieved in a significantly higher proportion treated with the Watchman FLX compared to the Amulet device. PDL was smaller with the FLX than the Amulet. Conceptual device design differences make interpretation of results complex, and additional studies with clinical outcomes are needed.

Comparison Between Echocardiography and Cardiac Computed Tomography in the Evaluation of Diastolic Dysfunction and Prediction of Heart Failure

Recent data indicate that left atrial (LA) function assessment by cardiac computed tomography (CT) is closely related to diastolic dysfunction (DD). Therefore, we aimed to perform a direct comparison between CT and echocardiography for diagnosis of advanced DD and prediction of future heart failure or cardiovascular death. We identified 340 patients who had both spiral cardiac CT and a proximate echocardiogram. LA total emptying fraction (LATEF), a measure of global LA function, was automatically calculated from CT data, as a surrogate for diastolic function and was compared with echocardiographic grades of diastolic function. The area under the receiver operating characteristic curve for LATEF to differentiate between advanced DD (grades 2 and 3) and all other grades was 0.84 (0.79 to 0.88). Over a median of 4 years, 69 events (admissions for heart failure and cardiovascular deaths) occurred. By multivariate Cox analysis, LATEF &lt;40% provided incremental prognostic information after adjustments for advanced DD by echocardiography (hazard ratio 2.15, 95% confidence interval 1.13 to 3.94). There was a significant interaction (p=0.03) between LATEF and echocardiography-based diastolic grades. Stratified analyses within the diastolic function groups revealed that LATEF &lt;40% was equivalent to echocardiography in predicting events in the subgroup with advanced DD by echocardiography (p=0.20) but was associated with a significantly higher event rates in patients with normal filling pressures (p=0.0001) or indeterminate diastolic function (p=0.04) by echocardiography. In conclusion, LA function derived from CT can accurately detect advanced DD diagnosed by echocardiography and has additive value to echocardiography-derived DD.

479 Commissural Misalignment Is An Independent Predictor Of Leaflet Thrombosis After Transcatheter Aortic Valve Implantation

Introduction: Hypo-attenuated leaflet thickening (HALT) - a marker of subclinical leaflet thrombosis commonly detected by cardiac computed tomography (CT) after transcatheter aortic valve implantation (TAVI) - may influence valve durability and function. We hypothesize that altered flow patterns in the aortic root as a consequence of commissural misalignment may contribute to the occurrence of HALT. Therefore, the purpose of this project was to compare the commissural alignment of the native and prosthetic aortic valves in cardiac CT in subjects with and without HALT and thereby identify potential predictors for subclinical leaflet thrombosis after TAVI.

Machine Learning Approach for Cardiovascular Risk and Coronary Artery Calcification Score.

Coronary artery calcification (CAC) could assist in the discovery of new risk elements for coronary artery disorder. CAC evaluation, on the other hand, is difficult due to the wide range of CAC in the populations. As a reason, evaluating and analysing data among research have become complicated. In the Research of Inherited Risk Factors for Coronary Atherosclerosis, we used CAC information to test the effects of different analytical methodologies on the correlation with recognized cardiovascular risk elements in asymptomatic patients. Cardiac computed tomography (CT) is also seeing an increase in examinations, and machine learning (ML) could assist with the growing amount of extracted data. Furthermore, there are other sectors in cardiac CT where machine learning could be crucial, including coronary calcium scoring, perfusion, and CT angiography. The establishment of risk evaluation algorithms based on information from CAC utilizing machine learning could assist in the categorization of patients undergoing cardiovascular into distinct risk groups and effectively adapt their treatments to their unique situations. Our findings imply that for forecasting CVD occurrences in asymptomatic people, age-sex segmentation by CAC percentile rank is as effective as absolute CAC scoring. Longitudinal population-based investigations are currently underway and would offer further definitive findings. While machine learning is a strong technology with a lot of possibilities, its implementations in the domain of cardiac CAC are generally in the early stages of development and are not currently commonly accessible in medical practise because of the requirement for substantial verification. Enhanced machine learning will, however, have a significant effect on cardiovascular and coronary artery calcification in the upcoming years.

Cardiac computed tomography radiomics: a narrative review of current status and future directions.

In an era of profound growth of medical data and rapid development of advanced imaging modalities, precision medicine increasingly requires further expansion of what can be interpreted from medical images. However, the current interpretation of cardiac computed tomography (CT) images mainly depends on subjective and qualitative analysis. Radiomics uses advanced image analysis to extract numerous quantitative features from digital images that are unrecognizable to the naked eye. Visualization of these features can reveal underlying connections between image phenotyping and biological characteristics and support clinical outcomes. Although research into radiomics on cardiovascular disease began only recently, several studies have indicated its potential clinical value in assessing future cardiac risk and guiding prevention and management strategies. Our review aimed to summarize the current applications of cardiac CT radiomics in the cardiovascular field and discuss its advantages, challenges, and future directions. We searched for English-language articles published between January 2010 and August 2021 in the databases of PubMed, Embase, and Google Scholar. The keywords used in the search included computed tomography or CT, radiomics, cardiovascular or cardiac. The current applications of radiomics in cardiac CT were found to mainly involve research into coronary plaques, perivascular adipose tissue (PVAT), myocardial tissue, and intracardiac lesions. Related findings on cardiac CT radiomics suggested the technique can assist the identification of vulnerable plaques or patients, improve cardiac risk prediction and stratification, discriminate myocardial pathology and etiologies behind intracardiac lesions, and offer new perspective and development prospects to personalized cardiovascular medicine. Cardiac CT radiomics can gather additional disease-related information at a microstructural level and establish a link between imaging phenotyping and tissue pathology or biology alone. Therefore, cardiac CT radiomics has significant clinical implications, including a contribution to clinical decision-making. Along with advancements in cardiac CT imaging, cardiac CT radiomics is expected to provide more precise phenotyping of cardiovascular disease for patients and doctors, which can improve diagnostic, prognostic, and therapeutic decision making in the future.

FIRST CARDIAC COMPUTED TOMOGRAPHY TYPICAL RADIATION DOSE VALUES FROM A SINGLE CENTRE IN NIGERIA: A PILOT STUDY.

Concern regarding radiation dose associated with cardiac computed tomography (CT) still exists and requires the use of diagnostic reference levels for dose optimisation. Typical median doses were established for 58 consented adult patients from a single centre for coronary artery calcium score (CACS) as volume computed tomography dose index (CTDIvol): 5.9mGy; dose length product (DLP): 86.6mGy*cm and cardiac CT angiography (CCTA) as CTDIvol: 11.1mGy; DLP: 190.8mGy*cm. Wide radiation dose variability in terms of CTDIvol was noted for CACS: 2.7-15.9mGy, a 6-fold, whereas for CCTA it ranges from 3.8 to 52.8mGy, a 14-fold. The DLP values for CACS range from 33.2 to 344.2mGy*cm, which is 10-fold, whereas for CCTA it ranges from 32.8 to 834.9mGy*cm, a 25-fold. The typical values compared lower than the radiation dose from other countries; however, the wide variability in dose remains a call for concern.

0705 Association of high risk obstructive sleep apnea with atherosclerotic plaque, coronary stenosis and coronary artery calcium score in asymptomatic young and middle-aged adults in The Miami Heart (MiHeart) Study at Baptist Health South Florida

Abstract Introduction Obstructive Sleep Apnea (OSA) is associated with clinical cardiovascular disease (CVD). There are limited data evaluating association with subclinical CVD measured by cardiac computed tomography (CT). We hypothesized there would be significant association between high OSA risk and atherosclerotic plaque, coronary stenosis, and coronary artery calcium (CAC) score in Miami Heart Study cohort free of CVD. Methods Data from CVD free 2359 participants, age 40-65 years (May 2015-Sept 2018) from greater Miami were analyzed. Cardiac CT measured CAC, coronary plaque burden and stenosis. High OSA risk was defined as either OSA diagnosis and/or a high risk from Berlin questionnaire. Logistic regression examined association between high OSA risk (reference low OSA risk) and any plaque, coronary stenosis &amp;gt;50%, and CAC (&amp;gt; 0 and &amp;gt; 100 vs 0) in unadjusted models, and after accounting for age, sex, race/ethnicity (model 2), and BMI, diabetes, high cholesterol, and smoking (model 3). Results 800 (34%) participants had high OSA risk; were more likely to be male, Hispanic, with higher CVD risk factor burden compared to low OSA risk. High OSA risk had higher prevalence of any plaque (60% vs 44%), coronary stenosis ≥ 50% (8.3% vs 4.8%), CAC scores &amp;gt;0-99 (34% vs 26%), and CAC score ≥ 100 (16% vs 12%), compared to low OSA risk (all p &amp;lt; 0.05). High OSA risk was associated with higher odds of any plaque, coronary stenosis ≥ 50%, high-risk plaque features, CAC &amp;gt; 0 and CAC ≥ 100 in univariable models. When adjusting for age, sex and race/ethnicity, these patterns persisted, with 1.58 higher odds of any plaque (95% CI 1.31, 1.91), 1.54 higher odds of coronary stenosis ≥ 50% (95% CI 1.08, 2.20), 1.37 higher odds of having CAC &amp;gt; 0 (95% CI 1.13, 1.66). Associations became non-statistically significant when adjusting for CVD risk factors, except any plaque, which was independently associated after fully adjusted (OR 1.31 (1.05, 1.63). Conclusion Individuals with high risk for OSA, significantly prone for CVD risk factor, have higher likelihood for presence of any plaque, significant stenosis and CAC scores &amp;gt; 0, with these associations mostly mediated by CVD risk factors. Support (If Any) Baptist Health South Florida

The Role of Cardiac Computed Tomography in Heart Failure.

Cardiac computed tomography (CT) is becoming a more widely applied tool in the diagnosis and management of a variety of cardiovascular conditions, including heart failure. The aim of this narrative review is to examine the role of cardiac CT in patients with heart failure. Coronary computed tomographic angiography has robust diagnostic accuracy for ruling out coronary artery disease. These data are reflected in updated guidelines from major cardiology organizations. New roles for cardiac CT in myocardial imaging, perfusion scanning, and periprocedural planning, execution, and monitoring are being implemented. Cardiac CT is useful in ruling out coronary artery disease its diagnostic accuracy, accessibility, and safety. It is also intricately linked to invasive cardiac procedures that patients with heart failure routinely undergo.

Long-term remodelling and arrhythmogenicity after myocardial infarction using a novel image-based estimator: the Scar Maturation Score

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – EU funding. Main funding source(s): European Research Council Introduction Post-infarction scar remodelling may start and evolve at a different pace across patients, often requiring years until the affected tissue becomes arrhythmogenic. Current risk stratification strategies lack sensitivity and specificity, and therefore more knowledge is required (e.g. new biomarkers and insights about the appropriate time point to evaluate them). Purpose We introduce a model of scar maturation and a novel estimator of scar remodelling based on fully automated analysis of cardiac computed tomography (CT). Methods We retrospectively included patients with history of myocardial infarction and available cardiac CT at the scarring stage (i.e. &amp;gt;6 months post-infarction). A fully automated pipeline was introduced including LV wall segmentation, reformatting to short-axis view, and LV wall thickness (WT) computation. 4 biomarkers representing moderate to severe LV wall thinning were considered (i.e. WT &amp;lt; {5, 4, 3, 2} mm). A statistical approach based on Bayesian Gaussian Process regression was used to model the biomarkers’ evolution. The Scar Maturation Score, a subject-specific parameter encoding the maturation stage relatively to the group-wise model was introduced, and its relationship with sustained ventricular arrhythmia (VA) was analysed. Results A total of 428 patients were included (age: 73±8 years, 83% males, LVEF: 46±11%), comprising 58 with 2 CT studies available (median scan delay: 2.1 years). The model estimated a progressive LV wall thinning over the years following the infarction. Patients without VA were typically assigned to low Scar Maturation Scores (i.e. early maturation stage), while VA patients were frequently associated with higher Scores. Scar Maturation Score, LVEF, area of wall thinning, and scar age were all significantly associated with VA (p &amp;lt; 0.0001), the Scar Maturation Score showing the higher correlation ratio. Conclusion Fully-automated LV thickness measurements from CT allow for a plausible modelling of scar maturation that, together with the patient-specific Scar Maturation Score, may be useful to improve patient treatment (e.g. determine optimal time point for patient screening) and risk stratification.

Left atrial anatomical variations correlate with atrial fibrillation sources near the left atrial ridge

Abstract Funding Acknowledgements Type of funding sources: None. Background Anatomical variations and characteristics of the left atrium (LA) may have a previously undescribed effect on source locations in atrial fibrillation (AF). Electrographic flow (EGF) mapping is a novel method used to estimate cardiac action potential flow in the atria that can detect AF sources in patients with persistent AF. The EGF technology offers a revolutionary mapping possibility for AF, however, it is not commercially available yet. Purpose This is the first study aiming to investigate the relationship between anatomical characteristics of the LA and non-PV sources detected by EGF in patients with persistent AF. Methods We collected cardiac computed tomography (CT) and EGF data from patients who underwent radiofrequency catheter ablation (CA) for persistent AF. EGF mapping creates a spatial and temporal reconstruction of electric potentials derived from endocardial unipolar electrograms. By analyzing EGF maps obtained during CA procedures, we localized non-PV sources in the LA. Patients with individual stable sources with a source activity above threshold (leading source present in more than &amp;gt; 26% of the time) were classified as having an S-Type EGF signature with source-dependent AF. Patients with no stable active source pattern and no leading source with a source activity above threshold were characterized as having a C-Type EGF signature consistent with source-independent AF. We assessed LA anatomical characteristics including size, LAA length, LAA ostial diameter, trajectory of the left superior pulmonary vein (LSPV) and its relation to the LAA on cardiac CT scans. Abutting LAA-LSPV was defined as cases when the LSPV touched the posterior aspect of LAA, and the maximal distance between the two structures was less than 2 mm. Those cases where the distance between LAA and LSPV was more than 2 mm were defined as non-abutting LAA-LSPV. Results Thirty patients were included in this study (mean age 62.4±6.8 years):23 patients had an S-Type EGF signature (77%), and 7 patients had a C-Type EGF signature (23%). We identified 10 patients with AF sources near the LA ridge, while twenty patients had no leading source near the LA ridge. LA anatomical characteristics, LAA length and ostial diameter showed no correlation with the presence of a leading source near the ridge. We described 19 patients with abutting LAA-LSPV, and 11 patients with non-abutting LAA-LSPV. Three out of 19 patients presented with a leading source near ridge in the abutting LAA-LSPV group, while 7 out of 11 patients presented with a leading source near ridge in the non-abutting LAA-LSPV group (p=0.01). Conclusion Our data suggests that non-abutting LAA-LSPV is associated with the presence of AF sources near the LA ridge.

Temporal changes and clinical significance of peridevice leak following left atrial appendage occlusion with Amplatzer devices.

BackgroundThe natural history of peridevice leak (PDL) following left atrial appendage occlusion (LAAO) is unknown. This study sought to investigate changes of PDL from 2 until 12 months after LAAO, using cardiac computed tomography (CT), and to assess the potential association between persistent PDL and clinical outcomesMethodsSingle‐center observational study of Amplatzer LAAO implants between 2010 and 2017 (n = 206). Patients with 2 and 12 months cardiac CT were included in the study (n = 153). Images were blindly analyzed. PDL was characterized by frequency and size at the device disc, lobe, and left atrial appendage contrast patency. Patients were followed for the composite outcome of ischemic stroke, transient ischemic attack, systemic embolism, or all‐cause death. Median follow up from LAAO was 3.1 (2.3–4.3) years.ResultsContrast patency was present in 101 (66%) and 72 (47%) (p < 0.001) at 2 and 12 months, respectively. PDL was identified at the disc in 103 (67%) patients at 2 months versus 93 (61%) at 12 months (p = 0.08), and at the lobe in 29 (19%) at both time points. PDL area at the disc did not change significantly over time, ∆ area: −8.95 mm (95% confidence interval [CI]: −18.9; 1.01) p = 0.08. Permanent atrial fibrillation was independently associated with persistent PDL. Persistent versus no PDL was associated with a 62% worse clinical outcome, however not statistically significant, hazard ratio (HR): 1.62 (95% CI: 0.9–2.93), p = 0.11.ConclusionPersistent PDL was frequently observed following LAAO with Amplatzer devices. The PDL frequency and size appeared unchanged between 2 and 12 months. Persistent PDL was not significantly associated with worse clinical outcomes, yet this needs further delineation in future studies.

Classification of severe aortic stenosis and outcomes after aortic valve replacement

Aortic valve calcium scoring by cardiac computed tomographic (CT) has been recommended as an alternative to classify the AS (aortic stenosis) severity, but it is unclear that whether CT findings would have additional value to discriminate significant AS subtypes including high gradient severe AS, classic low-flow, low gradient (LF-LG) AS, paradoxical LF-LG AS, and moderate AS. In this study, we examined the preoperative clinical and cardiac CT findings of different subtypes of AS in patients with surgical aortic valve replacement (AVR) and evaluated the subtype classification as a factor affecting post-surgical outcomes. This study included 511 (66.9 ± 8.8 years, 55% men) consecutive patients with severe AS who underwent surgical AVR. Aortic valve area (AVA) was obtained by echocardiography (AVAecho) and by CT (AVACT) using each modalities measurement of the left ventricular outflow tract. Patients with AS were classified as (1) high-gradient severe (n = 438), (2) classic LF-LG (n = 18), and (3) paradoxical LF-LG (n = 55) based on echocardiography. In all patients, 455 (89.0%) patients were categorized as severe AS according to the AVACT. However, 56 patients were re-classified as moderate AS (43 [9.8%] high-gradient severe AS, 5 [27.8%] classic LF-LG AS, and 8 [14.5%] paradoxical LF-LG AS) by AVACT. The classic LF-LG AS group presented larger AVACT and aortic annulus than those in high-gradient severe AS group and one third of them had AVACT ≥ 1.2 cm2. After multivariable adjustment, old age (hazard ratio [HR], 1.04, P = 0.049), high B-type natriuretic peptide (BNP) (HR, 1.005; P < 0.001), preoperative atrial fibrillation (HR, 2.75; P = 0.003), classic LF-LG AS (HR, 5.53, P = 0.004), and small aortic annulus on CT (HR, 0.57; P = 0.002) were independently associated with major adverse cardiac and cerebrovascular events (MACCE) after surgical AVR.

Machine Learning for the Prevalence and Severity of Coronary Artery Calcification in Nondialysis Chronic Kidney Disease Patients: A Chinese Large Cohort Study.

Purpose:This study sought to determine whether machine learning (ML) can be used to better identify the risk factors and establish the prediction models for the prevalence and severity of coronary artery calcification (CAC) in nondialysis chronic kidney disease (CKD) patients and compare the performance of distinctive ML models with conventional logistic regression (LR) model.Materials and Methods:In all, 3701 Chinese nondialysis CKD patients undergoing noncontrast cardiac computed tomography (CT) scanning were enrolled from November 2013 to December 2017. CAC score derived from the cardiac CT was calculated with the calcium scoring software and was used to assess and stratify the prevalence and severity of CAC. Four ML models (LR, random forest, support vector machine, and k-nearest neighbor) and the corresponding feature ranks were conducted. The model that incorporated the independent predictors was shown as the receiver-operating characteristic (ROC) curve. Area under the curve (AUC) was used to present the prediction value. ML model performance was compared with the traditional LR model using pairwise comparisons of AUCs.Results:Of the 3701 patients, 943 (25.5%) patients had CAC. Of the 943 patients with CAC, 764 patients (20.6%) and 179 patients (4.8%) had an Agatston CAC score of 1 to 300 and ≥300, respectively. The primary cohort and the independent validation cohort comprised 2957 patients and 744 patients, respectively. For the prevalence of CAC, the AUCs of ML models were from 0.78 to 0.82 in the training data set and the internal validation cohort. For the severity of CAC, the AUCs of the 4 ML models were from 0.67 to 0.70 in the training data set and from 0.53 to 0.70 in the internal validation cohort. For the prevalence of CAC, the AUC was 0.80 (95% confidence interval [CI]: 0.77-0.83) for ML (LR) versus 0.80 (95% CI: 0.77-0.83) for the traditional LR model (P=0.2533). For the severity of CAC, the AUC was 0.70 (95% CI: 0.63-0.77) for ML (LR) versus 0.70 (95% CI: 0.63-0.77) for traditional LR model (P=0.982).Conclusions:This study constructed prediction models for the presence and severity of CAC based on Agatston scores derived from noncontrast cardiac CT scanning in nondialysis CKD patients using ML, and showed ML LR had the best performance.

ECG and CT for the detection of left atrial enlargement in hypertensive individuals-a population-based study.

Left atrial enlargement (LAE) is associated with hypertension and an increased risk of cardiovascular morbidity and mortality. Guidelines for hypertension recommend LAE evaluation. We aimed to estimate the agreement of LAE as assessed by 12-lead electrocardiogram (ECG) and cardiac computed tomography (CT) in both the general population and hypertensive individuals. Cardiac CT and ECG were used to evaluate the presence of LAE in participants in the Copenhagen General Population Study. LAE, is defined as an LA volume above the 97.5% upper confidence limit by cardiac CT, as compared with multiple ECG criteria for LAE. A total of 3507 participants (47% males, age: 60 ± 10 years) were included. The prevalence of CT-defined LAE was 5.9% in the total population and 8.7% in participants with hypertension. In hypertensive individuals, LAE was identified by CT or by ECG in 31% with only a 4% overlap. ECG signs for anatomical LAE by CT had high negative predictive values between 93 and 96% but low sensitivity and positive predictive values. Specificity ranged from 27 to 93%. P-wave duration >120 ms was the best performing criterion, with a sensitivity of 48%, a specificity of 78%, and the highest area under the curve (0.66). We found a discrepancy in LAE prevalence when participants were assessed by CT and ECG, indicating that the two diagnostic modalities reflect different phenotypes of left atrial alterations. The diagnostic performance of ECG criteria for identifying anatomical LAE was poor.

Semi-automated three-dimensional segmentation for cardiac CT images using deep learning and randomly distributed points.

Given the prevalence of cardiovascular diseases (CVDs), the segmentation of the heart on cardiac computed tomography (CT) remains of great importance. Manual segmentation is time-consuming and intra-and inter-observer variabilities yield inconsistent and inaccurate results. Computer-assisted, and in particular, deep learning approaches to segmentation continue to potentially offer an accurate, efficient alternative to manual segmentation. However, fully automated methods for cardiac segmentation have yet to achieve accurate enough results to compete with expert segmentation. Thus, we focus on a semi-automated deep learning approach to cardiac segmentation that bridges the divide between a higher accuracy from manual segmentation and higher efficiency from fully automated methods. In this approach, we selected a fixed number of points along the surface of the cardiac region to mimic user interaction. Points-distance maps were then generated from these points selections, and a three-dimensional (3D) fully convolutional neural network (FCNN) was trained using points-distance maps to provide a segmentation prediction. Testing our method with different numbers of selected points, we achieved a Dice score from 0.742 to 0.917 across the four chambers. Specifically. Dice scores averaged 0.846 ± 0.059, 0.857 ± 0.052, 0.826 ± 0.062, and 0.824 ± 0.062 for the left atrium, left ventricle, right atrium, and right ventricle, respectively across all points selections. This point-guided, image-independent, deep learning segmentation approach illustrated a promising performance for chamber-by-chamber delineation of the heart in CT images.

Cardiac computed tomography-verified right ventricular lead position and outcomes in cardiac resynchronization therapy.

To evaluate the association between different right ventricular (RV) lead positions as assessed by cardiac computed tomography (CT) and echocardiographic and clinical outcomes in patients receiving cardiac resynchronization therapy (CRT). We reviewed patient records of all 278 patients included in two randomized controlled trials (ImagingCRT and ElectroCRT) for occurrence of heart failure (HF) hospitalization or all-cause death (primary endpoint) during long-term follow-up. Outcomes were compared between RV lead positions using adjusted Cox regression analysis. Six months after CRT implantation, we estimated left ventricular (LV) reverse remodeling by measuring LV end-systolic and end-diastolic volumes by echocardiography. Changes from baseline to 6months follow-up were compared between RV lead positions. Device-related complications were recorded at 6-month follow-up. During median (interquartile range) follow-up of 4.7 (2.9-7.1) years, the risk of meeting the primary endpoint was similar for patients with non-apical vs. apical RV lead position (adjusted hazard ratio (HR) 0.78, 95% confidence interval (CI) 0.54-1.12, p = 0.17) and free wall vs. septal RV lead position (adjusted HR 1.03, 95% CI 0.72-1.47, p = 0.86). Changes in LV ejection fraction and dimensions were similar with the different RV lead positions. We observed no differences in device-related complications relative to the RV lead position. In patients receiving CRT, the risk of HF hospitalization or all-cause death during long-term follow-up, and LV remodeling and incidence of device-related complications after 6months are not associated with different anatomical RV lead position as assessed by cardiac CT.