Quantification Of Coronary Artery Calcium Research Articles

Incidental detection and quantification of coronary artery calcification in ungated thoracic computed tomography for prognostic assessment. Preliminary results from the ICARUS registry

Abstract Background Coronary artery calcium (CAC) score improves cardiovascular risk stratification in asymptomatic individuals. However, whether CAC can be quantified in ungated thoracic computed tomography (uTCT) performed for other indications, and if this CAC score associates with cardiovascular outcomes, hasn’t been confirmed yet. Purpose We aim to analyse the extension of CAC and its prognostic value in terms of major adverse cardiovascular events (MACE) in patients who underwent uTCT for other indications. Methods We analysed a preliminary sample of 115 uTCT performed in 2015 in our hospital. We studied the presence and extension of CAC and quantified the global CAC score by Agatston method using validated software. Patients were categorized according to Coronary Artery Calcium Data and Reporting System (CAC-DRS) categories. Cardiovascular risk factors, lipidic control and pharmacological therapy of the cohort were registered. We analysed the time to first 3P-MACE, defined as cardiovascular death, non-fatal myocardial infarction and non-fatal stroke, whichever occurred first. A p-value <0.05 was considered statistically significant. Results Mean age of the cohort was 60.8±16.1 years (53% male, 53% smokers). Mean LDL-cholesterol was 121.67±34.91 mg/dL, and 35.1% received lipid-lowering therapy. In more than half (n=59, 51.3%) of the population CAC was detected. Mean CAC score was 312.99±774.48 Agatston units, and patients were categorized in CAC-DRS category 1 (n=19, 16.5%), category 2 (n=15, 13%) and category 3 (n=25, 21.7%). Patients with CAC were older (71±10.1 vs. 50.1±14.2 years, p<0.001), had higher burden of cardiovascular risk factors and prevalence of chronic kidney disease (CKD; n=8, 13.6% vs. n=1, 1.8%, p=0.03) and higher levels of triglycerides (143.2±80.5 vs. 106.8±36.2 mg/dL, p=0.004), but not LDL-cholesterol. Eleven (9.6%) patients presented 3P-MACE during a mean follow-up of 6.55±2.82 years. CAC score independently predicted 3P-MACE (HR 1.06 [1.01-1.11] per 100 Agatston units, p=0.02) along with CKD (HR 6.94 [1.58-30.46], p=0.01) and basal glucose levels (HR 1.03 [1.02-1.04] per mg/dL, p<0.001). Conclusions CAC can be incidentally detected in more than half of our unselected cohort of patients undergoing uTCT for other indications. CAC score quantification by Agatston method is feasible in uTCT and associates with a higher risk of MACE during follow-up.Incidental CAC detection in uTCT

Opportunistic screening of coronary artery calcification on non-gated conventional CT scans using artificial intelligence

Abstract Background Recent advancements in artificial intelligence (AI) led to the development of automatic Coronary artery calcification (CAC) analysis based on chest CT scans. The objective of this study is to assess the impact of AI-based CAC evaluation on improving the allocation of add-on therapies and further evaluation. Methods We used a novel propriety AI software to estimate CAC from non-gated, non-contrast chest CT scans. Patients were categorized into groups: low CAC 0-99 Agatston unit (AU), moderate CAC 100-399AU, and high CAC ≥ 400 AU. Exclusion criteria included prior myocardial infarction, percutaneous coronary intervention (PCI), coronary artery bypass graft (CABG), and life expectancy < 2 years. Patients classified as high CAC were invited to a dedicated clinic to perform a risk factor and clinical assessment, initiate appropriate medication, and refer to further testing as needed. For low-moderate CAC, referring physicians were informed of CAC status, encouraging guideline-based optimal preventive management. Results 1272 eligible patients between January 1th, 2023 to February 29th, 2024 were evaluated for inclusion. 641/1272 (50.3%) were excluded. Out of 631 enrolled patients, 84 (13%) patients were classified as high CAC, 163 (26%) as moderate and 348 (61%) as low CAC. At the dedicated clinic, appropriate low-density lipoprotein (LDL) goals were assigned and statins were initiated or dose adjusted where necessary. Twenty patients were referred to myocardial perfusion imaging, of which 2 were referred for invasive coronary angiography, one of them underwent PCI (Figure), and the other was referred for CABG. Conclusion This ongoing study indicates routine CAC quantification using AI software on chest CT scans can identify patients who may benefit from preventive cardiology services and lead them to appropriate cardiac treatment pathway. Further follow-up is required to assess its clinical impact.

A Template For Quality Assurance In AI-enabled Algorithms For Coronary Artery Calcium Quantification In Chest Ct: A Comparison With Subjective Human Scoring

A Template For Quality Assurance In AI-enabled Algorithms For Coronary Artery Calcium Quantification In Chest Ct: A Comparison With Subjective Human Scoring

Coronary artery calcium quantification technique using dual energy material decomposition: a simulation study

Coronary artery calcification is a significant predictor of cardiovascular disease, with current detection methods like Agatston scoring having limitations in sensitivity. This study aimed to evaluate the effectiveness of a novel CAC quantification method using dual-energy material decomposition, particularly its ability to detect low-density calcium and microcalcifications. A simulation study was conducted comparing the dual-energy material decomposition technique against the established Agatston scoring method and the newer volume fraction calcium mass technique. Detection accuracy and calcium mass measurement were the primary evaluation metrics. The dual-energy material decomposition technique demonstrated fewer false negatives than both Agatston scoring and volume fraction calcium mass, indicating higher sensitivity. In low-density phantom measurements, material decomposition resulted in only 7.41% false-negative (CAC = 0) measurements compared to 83.95% for Agatston scoring. For high-density phantoms, false negatives were removed (0.0%) compared to 20.99% in Agatston scoring. The dual-energy material decomposition technique presents a more sensitive and reliable method for CAC quantification.

Feasibility of virtual non-iodine coronary calcium scoring on dual source photon-counting coronary CT angiography: a dynamic phantom study

BackgroundThe aim of our current systematic dynamic phantom study was first, to optimize reconstruction parameters of coronary CTA (CCTA) acquired on photon counting CT (PCCT) for coronary artery calcium (CAC) scoring, and second, to assess the feasibility of calculating CAC scores from CCTA, in comparison to reference calcium scoring CT (CSCT) scans.MethodsIn this phantom study, an artificial coronary artery was translated at velocities corresponding to 0, < 60, and 60–75 beats per minute (bpm) within an anthropomorphic phantom. The density of calcifications was 100 (very low), 200 (low), 400 (medium), and 800 (high) mgHA/cm3, respectively. CCTA was reconstructed with the following parameters: virtual non-iodine (VNI), with and without iterative reconstruction (QIR level 2, QIR off, respectively); kernels Qr36 and Qr44f; slice thickness/increment 3.0/1.5 mm and 0.4/0.2 mm. The agreement in risk group classification between CACCCTA and CACCSCT scoring was measured using Cohen weighted linear κ with 95% CI.ResultsFor CCTA reconstructed with 0.4 mm slice thickness, calcium detectability was perfect (100%). At < 60 bpm, CACCCTA of low, and medium density calcification was underestimated by 53%, and 15%, respectively. However, CACCCTA was not significantly different from CACCSCT of very low, and high-density calcifications. The best risk agreement was achieved when CCTA was reconstructed with QIR off, Qr44f, and 0.4 mm slice thickness (κ = 0.762, 95% CI 0.671–0.853).ConclusionIn this dynamic phantom study, the detection of calcifications with different densities was excellent with CCTA on PCCT using thin-slice VNI reconstruction. Agatston scores were underestimated compared to CSCT but agreement in risk classification was substantial.Clinical relevance statementPhoton counting CT may enable the implementation of coronary artery calcium scoring from coronary CTA in daily clinical practice.Key PointsPhoton-counting CTA allows for excellent detectability of low-density calcifications at all heart rates.Coronary artery calcium scoring from coronary CTA acquired on photon counting CT is feasible, although improvement is needed.Adoption of the standard acquisition and reconstruction protocol for calcium scoring is needed for improved quantification of coronary artery calcium to fully employ the potential of photon counting CT.

A fully automated deep learning approach for coronary artery segmentation and comprehensive characterization.

Coronary computed tomography angiography (CCTA) allows detailed assessment of early markers associated with coronary artery disease (CAD), such as coronary artery calcium (CAC) and tortuosity (CorT). However, their analysis can be time-demanding and biased. We present a fully automated pipeline that performs (i) coronary artery segmentation and (ii) CAC and CorT objective analysis. Our method exploits supervised learning for the segmentation of the lumen, and then, CAC and CorT are automatically quantified. 281 manually annotated CCTA images were used to train a two-stage U-Net-based architecture. The first stage employed a 2.5D U-Net trained on axial, coronal, and sagittal slices for preliminary segmentation, while the second stage utilized a multichannel 3D U-Net for refinement. Then, a geometric post-processing was implemented: vessel centerlines were extracted, and tortuosity score was quantified as the count of branches with three or more bends with change in direction forming an angle >45°. CAC scoring relied on image attenuation. CAC was detected by setting a patient specific threshold, then a region growing algorithm was applied for refinement. The application of the complete pipeline required <5 min per patient. The model trained for coronary segmentation yielded a Dice score of 0.896 and a mean surface distance of 1.027 mm compared to the reference ground truth. Tracts that presented stenosis were correctly segmented. The vessel tortuosity significantly increased locally, moving from proximal, to distal regions (p < 0.001). Calcium volume score exhibited an opposite trend (p < 0.001), with larger plaques in the proximal regions. Volume score was lower in patients with a higher tortuosity score (p < 0.001). Our results suggest a linked negative correlation between tortuosity and calcific plaque formation. We implemented a fast and objective tool, suitable for population studies, that can help clinician in the quantification of CAC and various coronary morphological parameters, which is helpful for CAD risk assessment.

Artificial Intelligence in Coronary Artery Calcium Scoring Detection and Quantification.

Coronary artery calcium (CAC) is a marker of coronary atherosclerosis, and the presence and severity of CAC have been shown to be powerful predictors of future cardiovascular events. Due to its value in risk discrimination and reclassification beyond traditional risk factors, CAC has been supported by recent guidelines, particularly for the purposes of informing shared decision-making regarding the use of preventive therapies. In addition to dedicated ECG-gated CAC scans, the presence and severity of CAC can also be accurately estimated on non-contrast chest computed tomography scans performed for other clinical indications. However, the presence of such "incidental" CAC is rarely reported. Advances in artificial intelligence have now enabled automatic CAC scoring for both cardiac and non-cardiac CT scans. Various AI approaches, from rule-based models to machine learning algorithms and deep learning, have been applied to automate CAC scoring. Convolutional neural networks, a deep learning technique, have had the most successful approach, with high agreement with manual scoring demonstrated in multiple studies. Such automated CAC measurements may enable wider and more accurate detection of CAC from non-gated CT studies, thus improving the efficiency of healthcare systems to identify and treat previously undiagnosed coronary artery disease.

Fully automated coronary artery calcium quantification on electrocardiogram-gated non-contrast cardiac computed tomography using deep-learning with novel Heart-labelling method.

To develop an artificial intelligence (AI)-model which enables fully automated accurate quantification of coronary artery calcium (CAC), using deep learning (DL) on electrocardiogram (ECG)-gated non-contrast cardiac computed tomography (gated CCT) images. Retrospectively, 560 gated CCT images (including 60 synthetic images) performed at our institution were used to train AI-model, which can automatically divide heart region into five areas belonging to left main (LM), left anterior descending (LAD), circumflex (LCX), right coronary artery (RCA), and another. Total and vessel-specific CAC score (CACS) in each scan were manually evaluated. AI-model was trained with novel Heart-labelling method via DL according to the manual-derived results. Then, another 409 gated CCT images obtained in our institution were used for model validation. The performance of present AI-model was tested using another external cohort of 400 gated CCT images of Stanford Center for Artificial Intelligence of Medical Imaging by comparing with the ground truth. The overall accuracy of the AI-model for total CACS classification was excellent with Cohen's kappa of k = 0.89 and 0.95 (validation and test, respectively), which surpasses previous research of k = 0.89. Bland-Altman analysis showed little difference in individual total and vessel-specific CACS between AI-derived CACS and ground truth in test cohort (mean difference [95% confidence interval] were 1.5 [-42.6, 45.6], -1.5 [-100.5, 97.5], 6.6 [-60.2, 73.5], 0.96 [-59.2, 61.1], and 7.6 [-134.1, 149.2] for LM, LAD, LCX, RCA, and total CACS, respectively). Present Heart-labelling method provides a further improvement in fully automated, total, and vessel-specific CAC quantification on gated CCT.

Impaired metabolism predicts coronary artery calcification in women with systemic lupus erythematosus

Patients with systemic lupus erythematosus (SLE) exhibit a high risk for cardiovascular diseases (CVD) which is not fully explained by the classical Framingham risk factors. SLE is characterized by major metabolic alterations which can contribute to the elevated prevalence of CVD. A comprehensive analysis of the circulating metabolome and lipidome was conducted in a large cohort of 211 women with SLE who underwent a multi-detector computed tomography scan for quantification of coronary artery calcium (CAC), a robust predictor of coronary heart disease (CHD). Beyond traditional risk factors, including age and hypertension, disease activity and duration were independent risk factors for developing CAC in women with SLE. The presence of coronary calcium was associated with major alterations of circulating lipidome dominated by an elevated abundance of ceramides with very long chain fatty acids. Alterations in multiple metabolic pathways, including purine, arginine and proline metabolism, and microbiota-derived metabolites, were also associated with CAC in women with SLE. Logistic regression with bootstrapping of lipidomic and metabolomic variables were used to develop prognostic scores. Strikingly, combining metabolic and lipidomic variables with clinical and biological parameters markedly improved the prediction (area under the curve: 0.887, p<0.001) of the presence of coronary calcium in women with SLE. The present study uncovers the contribution of disturbed metabolism to the presence of coronary artery calcium and the associated risk of CHD in SLE. Identification of novel lipid and metabolite biomarkers may help stratifying patients for reducing CVD morbidity and mortality in SLE. INSERM and Sorbonne Université.

Use of Coronary Artery Calcium Quantification and Distribution for Coronary Vascular Disease Risk Reclassification in a Primary Prevention Setting

In a large screening program of asymptomatic middle-aged individuals, we sought to assess the degree of risk reclassification provided by comparing multiethnic study on subclinical atherosclerosis coronary artery calcium scoring (CACS) versus atherosclerotic cardiovascular disease (ASCVD) and Reynolds risk score (RRS) score. All 5,324 consecutive patients (aged 57 ± 8years, 76% male) who underwent CACS screening at the Cleveland Clinic as part of a primary prevention executive health between March 16 and October 21 were included. The 10-year ASCVD, RRS, and multiethnic study on subclinical atherosclerosis CACS (MESA-CACS) risk scores were calculated and categorized as <1, 1 to 4.99, 5 to 9.99, and ≥10%. Compared with ASCVD, using MESA-CACS resulted in a downgraded risk in 1,667 subjects (31%), whereas 738 (14%) had an upgrade in risk (total of 45% reclassification). Similarly, compared with RRS, using MESA-CACS resulted in an upgraded risk in 797 (15%) and a downgrade in 1,380 (26%) subjects (total of 41% reclassification). However, by further dividing by the distribution of the coronary calcification, ASCVD overestimates the risk only for patients with coronary artery calcium (CAC) in 0 or 1 coronary artery only, whereas MESA-CACS overestimates if the CAC was noted in ≥2 arteries. Similarly, RRS only overestimates the risk for patients with 0 CAC, whereas it underestimates the risk for patients with any CAC. In conclusion, the use of MESA-CACS, along with CAC distribution in primary prevention clinics, results in differential and significant reclassification of traditional scores when calculating the 10-years coronary vascular disease risk. Overall, RRS underestimates and ASCVD overestimates the cardiovascular disease risk compared with MESA-CACS.

Impact of Cardiac Motion on coronary artery calcium scoring using a virtual non-iodine algorithm on photon-counting detector CT: a dynamic phantom study.

This study assessed the impact of cardiac motion and in-vessel attenuation on coronary artery calcium (CAC) scoring using virtual non-iodine (VNI) against virtual non-contrast (VNC) reconstructions on photon-counting detector CT. Two artificial vessels containing calcifications and different in-vessel attenuations (500, 800HU) were scanned without (static) and with cardiac motion (60, 80, 100 beats per minute [bpm]). Images were post-processed using a VNC and VNI algorithm at 70keV and quantum iterative reconstruction (QIR) strength 2. Calcium mass, Agatston scores, cardiac motion susceptibility (CMS)-indices were compared to physical mass, static scores as well as between reconstructions, heart rates and in-vessel attenuations. VNI scores decreased with rising heart rate (p < 0.01) and showed less underestimation than VNC scores (p < 0.001). Only VNI scores were similar to the physical mass at static measurements, and to static scores at 60bpm. Agatston scores using VNI were similar to static scores at 60 and 80bpm. Standard deviation of CMS-indices was lower for VNI-based than for VNC-based CAC scoring. VNI scores were higher at 500 than 800HU (p < 0.001) and higher than VNC scores (p < 0.001) with VNI scores at 500 HU showing the lowest deviation from the physical reference. VNI-based CAC quantification is influenced by cardiac motion and in-vessel attenuation, but least when measuring Agatston scores, where it outperforms VNC-based CAC scoring.

Ex vivo coronary calcium volume quantification using a high-spatial-resolution clinical photon-counting-detector computed tomography.

Coronary artery calcification (CAC) is an important indicator of coronary disease. Accurate volume quantification of CAC is challenging using computed tomography (CT) due to calcium blooming, which is a consequence of limited spatial resolution. Ex vivo coronary specimens were scanned on an ultra-high-resolution (UHR) clinical photon-counting detector (PCD) CT scanner, and the accuracy of CAC volume estimation was compared with a state-of-the-art conventional energy-integrating detector (EID) CT, a previous-generation investigational PCD-CT, and micro-CT. CAC specimens () were scanned on EID-CT and PCD-CT using matched parameters (120kV, 9.3mGy ). EID-CT images were reconstructed using our institutional routine clinical protocol for CAC quantification. UHR PCD-CT data were reconstructed using a sharper kernel. An image-based denoising algorithm was applied to the PCD-CT images to achieve similar noise levels as EID-CT. Micro-CT images served as the volume reference standard. Calcification images were segmented, and their volume estimates were compared. The CT data were further compared with previous work using an investigational PCD-CT. Compared with micro-CT, CT volume estimates had a mean absolute percent error of for clinical PCD-CT, for EID-CT, and for previous-generation PCD-CT. Clinical PCD-CT absolute percent error was significantly () lower than both EID-CT and previous generation PCD-CT. The mean calcification CT number and contrast-to-noise ratio were both significantly () higher in clinical PCD-CT relative to EID-CT. UHR clinical PCD-CT showed reduced calcium blooming artifacts and further enabled improved accuracy of CAC quantification beyond that of conventional EID-CT and previous generation PCD-CT systems.

Cardiac Virtual Noncontrast Images for Calcium Quantification with Photon-counting Detector CT

Purpose To assess the accuracy of aortic valve calcium (AVC), mitral annular calcium (MAC), and coronary artery calcium (CAC) quantification and risk stratification using virtual noncontrast (VNC) images from late enhancement photon-counting detector CT as compared with true noncontrast images. Materials and Methods This retrospective, institutional review board–approved study evaluated patients undergoing photon-counting detector CT between January and September 2022. VNC images were reconstructed from late enhancement cardiac scans at 60, 70, 80, and 90 keV using quantum iterative reconstruction (QIR) strengths of 2–4. AVC, MAC, and CAC were quantified on VNC images and compared with quantification of AVC, MAC, and CAC on true noncontrast images using Bland-Altman analyses, regression models, intraclass correlation coefficients (ICC), and Wilcoxon tests. Agreement between severe aortic stenosis likelihood categories and CAC risk categories determined from VNC and true noncontrast images was assessed by weighted κ analysis. Results Ninety patients were included (mean age, 80 years ± 8 [SD]; 49 male patients). Scores were similar on true noncontrast images and VNC images at 80 keV for AVC and MAC, regardless of QIR strengths, and VNC images at 70 keV with QIR 4 for CAC (all P > .05). The best results were achieved using VNC images at 80 keV with QIR 4 for AVC (mean difference, 3; ICC = 0.992; r = 0.98) and MAC (mean difference, 6; ICC = 0.998; r = 0.99), and VNC images at 70 keV with QIR 4 for CAC (mean difference, 28; ICC = 0.996; r = 0.99). Agreement between calcification categories was excellent on VNC images at 80 keV for AVC (κ = 0.974) and on VNC images at 70 keV for CAC (κ = 0.967). Conclusion VNC images from cardiac photon-counting detector CT enables patient risk stratification and accurate quantification of AVC, MAC, and CAC. Keywords: Coronary Arteries, Aortic Valve, Mitral Valve, Aortic Stenosis, Calcifications, Photon-counting Detector CT Supplemental material is available for this article © RSNA, 2023

Automated cardiovascular risk categorization through AI-driven coronary calcium quantification in cardiac PET acquired attenuation correction CT

BackgroundWe present an automatic method for coronary artery calcium (CAC) quantification and cardiovascular risk categorization in CT attenuation correction (CTAC) scans acquired at rest and stress during cardiac PET/CT. The method segments CAC according to visual assessment rather than the commonly used CT-number threshold. MethodsThe method decomposes an image containing CAC into a synthetic image without CAC and an image showing only CAC. Extensive evaluation was performed in a set of 98 patients, each having rest and stress CTAC scans and a dedicated calcium scoring CT (CSCT). Standard manual calcium scoring in CSCT provided the reference standard. ResultsThe interscan reproducibility of CAC quantification computed as average absolute relative differences between CTAC and CSCT scan pairs was 75% and 85% at rest and stress using the automatic method compared to 121% and 114% using clinical calcium scoring. Agreement between automatic risk assessment in CTAC and clinical risk categorization in CSCT resulted in linearly weighted kappa of 0.65 compared to 0.40 between CTAC and CSCT using clinically used calcium scoring. ConclusionThe increased interscan reproducibility achieved by our method may allow routine cardiovascular risk assessment in CTAC, potentially relieving the need for dedicated CSCT.

Coronary artery calcium quantification: comparison between filtered-back projection, hybrid iterative reconstruction, and deep learning reconstruction techniques.

The reference protocol for the quantification of coronary artery calcium (CAC) should be updated to meet the standards of modern imaging techniques. To assess the influence of filtered-back projection (FBP), hybrid iterative reconstruction (IR), and three levels of deep learning reconstruction (DLR) on CAC quantification on both in vitro and in vivo studies. In vitro study was performed with a multipurpose anthropomorphic chest phantom and small pieces of bones. The real volume of each piece was measured using the water displacement method. In the in vivo study, 100 patients (84 men; mean age = 71.2 ± 8.7 years) underwent CAC scoring with a tube voltage of 120 kVp and image thickness of 3 mm. The image reconstruction was done with FBP, hybrid IR, and three levels of DLR including mild (DLRmild), standard (DLRstd), and strong (DLRstr). In the in vitro study, the calcium volume was equivalent (P = 0.949) among FBP, hybrid IR, DLRmild, DLRstd, and DLRstr. In the in vivo study, the image noise was significantly lower in images that used DLRstr-based reconstruction, when compared images other reconstructions (P < 0.001). There were no significant differences in the calcium volume (P = 0.987) and Agatston score (P = 0.991) among FBP, hybrid IR, DLRmild, DLRstd, and DLRstr. The highest overall agreement of Agatston scores was found in the DLR groups (98%) and hybrid IR (95%) when compared to standard FBP reconstruction. The DLRstr presented the lowest bias of agreement in the Agatston scores and is recommended for the accurate quantification of CAC.

Deep Learning-Based Automated Quantification of Coronary Artery Calcification for Contrast-Enhanced Coronary Computed Tomographic Angiography.

We evaluated the accuracy of a deep learning-based automated quantification algorithm for coronary artery calcium (CAC) based on enhanced ECG-gated coronary CT angiography (CCTA) with dedicated coronary calcium scoring CT (CSCT) as the reference. This retrospective study included 315 patients who underwent CSCT and CCTA on the same day, with 200 in the internal and 115 in the external validation sets. The calcium volume and Agatston scores were calculated using both the automated algorithm in CCTA and the conventional method in CSCT. The time required for computing calcium scores using the automated algorithm was also evaluated. Our automated algorithm extracted CACs in less than five minutes on average with a failure rate of 1.3%. The volume and Agatston scores by the model showed high agreement with those from CSCT with concordance correlation coefficients of 0.90-0.97 for the internal and 0.76-0.94 for the external. The accuracy for classification was 92% with a 0.94 weighted kappa for the internal and 86% with a 0.91 weighted kappa for the external set. The deep learning-based and fully automated algorithm efficiently extracted CACs from CCTA and reliably assigned categorical classification for Agatston scores without additional radiation exposure.

An automated segmentation of coronary artery calcification using deep learning in specific region limitation.

Coronary artery calcification (CAC) is a frequent disease of the arteries that supply the surface of the heart muscle. Leaving a severe disease untreated can make it permanent. Computer tomography (CT), which is well known for its ability to quantify the Agatston score, is used to visualize high-resolution CACs. CAC segmentation is still an important topic. Our goal is to automatically segment CAC in a specific area and measure the Agatston score in 2D images. The heart region is limited using a threshold, unused structures are removed using 2D connectivity (muscle, lung, ribcage), the heart cavity is extracted using the convex hull of the lungs, and the CAC is then segmented in 2D using a convolutional neural network (U-Net models/SegNet-VGG16 with transfer learning). The Agatston score prediction is calculated for CAC quantification. The proposed strategy is tested through experiments, which yield encouraging outcomes. Graphical Abstract Deep learning for CAC segmentation in CT images.

Association Between Measures of Body Composition and Coronary Calcium: Findings From the Multi-Ethnic Study of Atherosclerosis.

Background Obesity, as measured by body mass index, is widely recognized as a risk factor for the development of cardiovascular disease. However, the role of body composition components such as fat and lean mass is not well studied. Methods and Results A total of 3129 patients who underwent computed tomography scans for quantification of coronary artery calcification and had bioelectrical impedance analysis of body composition (fat mass and fat-free mass) during exam 5 of MESA (Multi-Ethnic Study of Atherosclerosis) were included in this cross-sectional analysis. Multivariable adjusted linear regression analysis was performed to assess the relationship between both fat mass and fat-free mass to prevalent coronary artery calcification, a marker of subclinical coronary artery disease quantified by both the coronary artery calcification (CAC) Agatston score and the spatially weighted calcium score. CAC and spatially weighted calcium score were natural log-transformed for analysis as continuous variables. Fat-free mass, but not fat mass, was independently associated with CAC. There was a 7.6% prevalence risk difference for CAC>0 per 10 kg. Fat-free mass was also significantly associated with natural log of CAC (coefficient=0.272, P<0.001). Both fat-free mass and fat mass were positively associated with natural log of spatially weighted calcium score, with risk difference coefficients of 0.729 and 0.359, respectively (P<0.001). Conclusions In this cross-sectional study, higher lean mass by bioelectrical impedance analysis and, to a lesser extent, higher fat mass by bioelectrical impedance analysis were significantly associated with higher coronary calcium, a marker of subclinical cardiovascular disease. Further exploration of the relationship between components of body composition and the development of cardiovascular disease is warranted.

Impact of deep learning image reconstructions (DLIR) on coronary artery calcium quantification

BackgroundDeep learning image reconstructions (DLIR) have been recently introduced as an alternative to filtered back projection (FBP) and iterative reconstruction (IR) algorithms for computed tomography (CT) image reconstruction. The aim of this study was to evaluate the effect of DLIR on image quality and quantification of coronary artery calcium (CAC) in comparison to FBP.MethodsOne hundred patients were consecutively enrolled. Image quality–associated variables (noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR)) as well as CAC-derived parameters (Agatston score, mass, and volume) were calculated from images reconstructed by using FBP and three different strengths of DLIR (low (DLIR_L), medium (DLIR_M), and high (DLIR_H)). Patients were stratified into 4 risk categories according to the Coronary Artery Calcium - Data and Reporting System (CAC-DRS) classification: 0 Agatston score (very low risk), 1–99 Agatston score (mildly increased risk), Agatston 100–299 (moderately increased risk), and ≥ 300 Agatston score (moderately-to-severely increased risk).ResultsIn comparison to standard FBP, increasing strength of DLIR was associated with a significant and progressive decrease of image noise (p < 0.001) alongside a significant and progressive increase of both SNR and CNR (p < 0.001). The use of incremental levels of DLIR was associated with a significant decrease of Agatston CAC score and CAC volume (p < 0.001), while mass score remained unchanged when compared to FBP (p = 0.232). The underestimation of Agatston CAC led to a CAC-DRS misclassification rate of 8%.ConclusionDLIR systematically underestimates Agatston CAC score. Therefore, DLIR should be used cautiously for cardiovascular risk assessment.Key Points• In coronary artery calcium imaging, the implementation of deep learning image reconstructions improves image quality, by decreasing the level of image noise.• Deep learning image reconstructions systematically underestimate Agatston coronary artery calcium score.• Deep learning image reconstructions should be used cautiously in clinical routine to measure Agatston coronary artery calcium score for cardiovascular risk assessment.

Abstract 12604: Coronary Artery Calcium Score Determined From Pre-Ablation Pulmonary Vein Computed Tomography Allows for Diagnosis of Coronary Artery Disease and Improved Atherosclerotic Cardiovascular Disease Risk Prediction

Objective: Coronary artery calcification (CAC) is a useful adjunct for atherosclerotic cardiovascular disease (ASCVD) risk stratification and can influence preventive strategies. However, CAC screening is infrequently done, even among cardiology patients. Ancillary imaging, for example pulmonary vein CT (CTPV) performed prior to atrial arrhythmia ablation, may allow for CAC quantification. Our goal was to assess CAC prevalence on pre-ablation CTPVs and to assess how this information could influence risk stratification. Methods: Patients undergoing atrial arrhythmia ablation at NYU from 4/1/18 to 4/30/19 were identified and all available medical records reviewed. CTPVs were reviewed and CAC graded semi-quantitatively (0, 1-9, 10-99, 100-399, 400-999, 1000+). 10-yr ASCVD and MESA scores were calculated for patients without known ASCVD. Results: 739 patients underwent ablations and 660 had CTPV available for assessment. Mean age was 63.4 ± 11.4 years, 33.9% were female, and 17.1% had a history of ASCVD (prior MI, CVA/TIA, CAD, PAD). CAC was present in 62% of patients, including 57.2% without ASCVD history (22.3% of these had CAC ≥100). Complete data for the calculation of 10-year AHA/ACC and MESA risk scores were present in 27.9% and 32.4%, respectively. MESA CHD risk score allows for the incorporation of CAC for improved predictive capacity. Median MESA risk was 6.5% without and 6.7% with median CAC included. Incorporating CAC into MESA risk decreased risk in 110 patients (79 of these had CAC &lt;10), and increased risk in 102 patients (101 had CAC ≥10) ( Figure ). AHA/ACC ASCVD risk distribution is displayed in Figure. Guidelines recommend consideration of statins in those with ASCVD risk &lt;7.5% and non-zero CAC. 52.4% of patients with ASCVD risk &lt;7.5 had non-zero CAC ( Figure ) with 18.2% on a statin at the time of their procedure. Conclusion: Quantification of CAC on CTPV can provide a means to diagnose CAD, improve risk stratification and impact primary preventive treatment.