ABSTRACT
Purpose To use unsupervised machine learning to identify phenotypic clusters with increased risk of arrhythmic mitral valve prolapse (MVP). Materials and Methods This retrospective study included patients with MVP without hemodynamically significant mitral regurgitation or left ventricular (LV) dysfunction undergoing late gadolinium enhancement (LGE) cardiac MRI between October 2007 and June 2020 in 15 European tertiary centers. The study end point was a composite of sustained ventricular tachycardia, (aborted) sudden cardiac death, or unexplained syncope. Unsupervised data-driven hierarchical k-mean algorithm was utilized to identify phenotypic clusters. The association between clusters and the study end point was assessed by Cox proportional hazards model. Results A total of 474 patients (mean age, 47 years ± 16 [SD]; 244 female, 230 male) with two phenotypic clusters were identified. Patients in cluster 2 (199 of 474, 42%) had more severe mitral valve degeneration (ie, bileaflet MVP and leaflet displacement), left and right heart chamber remodeling, and myocardial fibrosis as assessed with LGE cardiac MRI than those in cluster 1. Demographic and clinical features (ie, symptoms, arrhythmias at Holter monitoring) had negligible contribution in differentiating the two clusters. Compared with cluster 1, the risk of developing the study end point over a median follow-up of 39 months was significantly higher in cluster 2 patients (hazard ratio: 3.79 [95% CI: 1.19, 12.12], P = .02) after adjustment for LGE extent. Conclusion Among patients with MVP without significant mitral regurgitation or LV dysfunction, unsupervised machine learning enabled the identification of two phenotypic clusters with distinct arrhythmic outcomes based primarily on cardiac MRI features. These results encourage the use of in-depth imaging-based phenotyping for implementing arrhythmic risk prediction in MVP. Keywords: MR Imaging, Cardiac, Cardiac MRI, Mitral Valve Prolapse, Cluster Analysis, Ventricular Arrhythmia, Sudden Cardiac Death, Unsupervised Machine Learning Supplemental material is available for this article. © RSNA, 2024.
Subject(s)
Mitral Valve Prolapse , Phenotype , Unsupervised Machine Learning , Humans , Mitral Valve Prolapse/diagnostic imaging , Female , Male , Middle Aged , Retrospective Studies , Registries , Magnetic Resonance Imaging, Cine/methods , Arrhythmias, Cardiac/diagnostic imaging , Arrhythmias, Cardiac/physiopathology , Adult , Magnetic Resonance ImagingABSTRACT
Background: The recent Cardiovascular Disease in Adolescents with Chronic Disease (CDACD) study showed enhanced aortic stiffness and wall thickness in adolescents with various chronic disorders. Enhanced aortic stiffness can increase left ventricular (LV) afterload and trigger a cascade of adverse arterioventricular interaction. Here, we investigate the relation between aortic changes and LV function in the CDACD study participants. Methods: This cross-sectional study included 114 adolescents 12-18 years old with cystic fibrosis (CF, n = 24), corrected coarctation of the aorta (CoA, n = 25), juvenile idiopathic arthritis (JIA, n = 20), obesity (n = 20), and healthy controls (n = 25). Aortic pulse wave velocity (PWV), which reflects aortic stiffness, and aortic wall thickness (AWT) were assessed with cardiovascular magnetic resonance imaging (CMR). Echocardiography was employed to study conventional markers of LV function, as well as LV global longitudinal strain (LVGLS), which is an established (pre)clinical marker of LV dysfunction. Results: First, aortic PWV and AWT were increased in all chronic disease groups, compared to controls. Second, in adolescents with CoA, JIA, and obesity, echocardiography showed a decreased LVGLS, while LV dimensions and conventional LV function markers were similar to controls. Third, multivariable linear regression identified aortic PWV as the most important determinant of their decreased LVGLS (standardized ß -0.522, p < 0.001). Conclusions: The decreased LVGLS in several adolescent chronic disease groups was associated with enhanced aortic PWV, which might reflect adverse arterioventricular interaction. Whether the decreased LVGLS in the chronic disease groups could negatively impact their long-term cardiovascular outcomes requires further study.
ABSTRACT
BACKGROUND: The 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. METHODS: In 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. RESULTS: For 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). CONCLUSION: In 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 STATEMENT: Photon counting CT may enable the implementation of coronary artery calcium scoring from coronary CTA in daily clinical practice. KEY POINTS: Photon-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.
ABSTRACT
In cancer research there is much interest in building and validating outcome prediction models to support treatment decisions. However, because most outcome prediction models are developed and validated without regard to the causal aspects of treatment decision making, many published outcome prediction models may cause harm when used for decision making, despite being found accurate in validation studies. Guidelines on prediction model validation and the checklist for risk model endorsement by the American Joint Committee on Cancer do not protect against prediction models that are accurate during development and validation but harmful when used for decision making. We explain why this is the case and how to build and validate models that are useful for decision making.
Subject(s)
Algorithms , Humans , Causality , Clinical Decision-Making , Neoplasms/therapy , Quality ImprovementABSTRACT
BACKGROUND: The absence of population-stratified cardiovascular magnetic resonance (CMR) reference ranges from large cohorts is a major shortcoming for clinical care. OBJECTIVES: This paper provides age-, sex-, and ethnicity-specific CMR reference ranges for atrial and ventricular metrics from the Healthy Hearts Consortium, an international collaborative comprising 9,088 CMR studies from verified healthy individuals, covering the complete adult age spectrum across both sexes, and with the highest ethnic diversity reported to date. METHODS: CMR studies were analyzed using certified software with batch processing capability (cvi42, version 5.14 prototype, Circle Cardiovascular Imaging) by 2 expert readers. Three segmentation methods (smooth, papillary, anatomic) were used to contour the endocardial and epicardial borders of the ventricles and atria from long- and short-axis cine series. Clinically established ventricular and atrial metrics were extracted and stratified by age, sex, and ethnicity. Variations by segmentation method, scanner vendor, and magnet strength were examined. Reference ranges are reported as 95% prediction intervals. RESULTS: The sample included 4,452 (49.0%) men and 4,636 (51.0%) women with average age of 61.1 ± 12.9 years (range: 18-83 years). Among these, 7,424 (81.7%) were from White, 510 (5.6%) South Asian, 478 (5.3%) mixed/other, 341 (3.7%) Black, and 335 (3.7%) Chinese ethnicities. Images were acquired using 1.5-T (n = 8,779; 96.6%) and 3.0-T (n = 309; 3.4%) scanners from Siemens (n = 8,299; 91.3%), Philips (n = 498; 5.5%), and GE (n = 291, 3.2%). CONCLUSIONS: This work represents a resource with healthy CMR-derived volumetric reference ranges ready for clinical implementation.
Subject(s)
Healthy Volunteers , Magnetic Resonance Imaging, Cine , Predictive Value of Tests , Humans , Middle Aged , Male , Female , Adult , Aged , Reference Values , Adolescent , Young Adult , Aged, 80 and over , Magnetic Resonance Imaging, Cine/standards , Sex Factors , Age Factors , Heart Atria/diagnostic imaging , Heart Ventricles/diagnostic imaging , Reproducibility of Results , Ethnicity , Ventricular Function, Left , Race FactorsABSTRACT
Background: Pulmonary artery (PA) strain is associated with structural and functional alterations of the vessel and is an independent predictor of cardiovascular events. The relationship of PA strain to metabolomics in participants without cardiovascular disease is unknown. Methods: In the current study, community-based older adults, without known cardiovascular disease, underwent simultaneous cine cardiovascular magnetic resonance (CMR) imaging, clinical examination, and serum sampling. PA global longitudinal strain (GLS) analysis was performed by tracking the change in distance from the PA bifurcation to the pulmonary annular centroid, using standard cine CMR images. Circulating metabolites were measured by cross-sectional targeted metabolomics analysis. Results: Among n = 170 adults (mean age 71 ± 6.3 years old; 79 women), mean values of PA GLS were 16.2 ± 4.4%. PA GLS was significantly associated with age (ß = -0.13, P = 0.017), heart rate (ß = -0.08, P = 0.001), dyslipidemia (ß = -2.37, P = 0.005), and cardiovascular risk factors (ß = -2.49, P = 0.001). Alanine (ß = -0.007, P = 0.01) and proline (ß = -0.0009, P = 0.042) were significantly associated with PA GLS after adjustment for clinical risk factors. Medium and long-chain acylcarnitines were significantly associated with PA GLS (C12, P = 0.027; C12-OH/C10-DC, P = 0.018; C14:2, P = 0.036; C14:1, P = 0.006; C14, P = 0.006; C14-OH/C12-DC, P = 0.027; C16:3, P = 0.019; C16:2, P = 0.006; C16:1, P = 0.001; C16:2-OH, P = 0.016; C16:1-OH/C14:1-DC, P = 0.028; C18:1-OH/C16:1-DC, P = 0.032). Conclusion: By conventional CMR, PA GLS was associated with aging and vascular risk factors among a contemporary cohort of older adults. Metabolic pathways involved in PA stiffness may include gluconeogenesis, collagen synthesis, and fatty acid oxidation.
ABSTRACT
Aging-induced aortic stiffness has been associated with altered fatty acid metabolism. We studied aortic stiffness using cardiac magnetic resonance (CMR)-assessed ventriculo-arterial coupling (VAC) and novel aortic (AO) global longitudinal strain (GLS) combined with targeted metabolomic profiling. Among community older adults without cardiovascular disease, VAC was calculated as aortic pulse wave velocity (PWV), a marker of arterial stiffness, divided by left ventricular (LV) GLS. AOGLS was the maximum absolute strain measured by tracking the phasic distance between brachiocephalic artery origin and aortic annulus. In 194 subjects (71 ± 8.6 years; 88 women), AOGLS (mean 5.6 ± 2.1%) was associated with PWV (R = -0.3644, p < 0.0001), LVGLS (R = 0.2756, p = 0.0001) and VAC (R = -0.3742, p <0.0001). Stiff aorta denoted by low AOGLS <4.26% (25th percentile) was associated with age (OR 1.13, 95% CI 1.04-1.24, p = 0.007), body mass index (OR 1.12, 95% CI 1.01-1.25, p = 0.03), heart rate (OR 1.04, 95% CI 1.01-1.06, p = 0.011) and metabolites of medium-chain fatty acid oxidation: C8 (OR 1.005, p = 0.026), C10 (OR 1.003, p = 0.036), C12 (OR 1.013, p = 0.028), C12:2-OH/C10:2-DC (OR 1.084, p = 0.032) and C16-OH (OR 0.82, p = 0.006). VAC was associated with changes in long-chain hydroxyl and dicarboxyl carnitines. Multivariable models that included acyl-carnitine metabolites, but not amino acids, significantly increased the discrimination over clinical risk factors for prediction of AOGLS (AUC [area-under-curve] 0.73 to 0.81, p = 0.037) and VAC (AUC 0.78 to 0.87, p = 0.0044). Low AO GLS and high VAC were associated with altered medium-chain and long-chain fatty acid oxidation, respectively, which may identify early metabolic perturbations in aging-associated aortic stiffening. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02791139.
ABSTRACT
PURPOSE: The aim of this study was to characterize a second-generation wide-detector dual-layer spectral computed tomography (CT) system for material quantification accuracy, acquisition parameter and patient size dependencies, and tissue characterization capabilities. METHODS: A phantom with multiple tissue-mimicking and material-specific inserts was scanned with a dual-layer spectral detector CT using different tube voltages, collimation widths, radiation dose levels, and size configurations. Accuracy of iodine density maps and virtual monoenergetic images (MonoE) were investigated. Additionally, differences between conventional and MonoE 70 keV images were calculated to evaluate acquisition parameter and patient size dependencies. To demonstrate material quantification and differentiation, liver-mimicking inserts with adipose and iron were analyzed with a two-base decomposition utilizing MonoE 50 and 150 keV, and root mean square error (RMSE) for adipose and iron content was reported. RESULTS: Measured inserts exhibited quantitative accuracy across a wide range of MonoE levels. MonoE 70 keV images demonstrated reduced dependence compared to conventional images for phantom size (1 vs. 27 HU) and acquisition parameters, particularly tube voltage (4 vs. 37 HU). Iodine density quantification was successful with errors ranging from -0.58 to 0.44 mg/mL. Similarly, inserts with different amounts of adipose and iron were differentiated, and the small deviation in values within inserts corresponded to a RMSE of 3.49 ± 1.76% and 1.67 ± 0.84 mg/mL for adipose and iron content, respectively. CONCLUSION: The second-generation dual-layer CT enables acquisition of quantitatively accurate spectral data without compromises from differences in patient size and acquisition parameters.
Subject(s)
Iodine , Tomography, X-Ray Computed , Humans , Signal-To-Noise Ratio , Tomography, X-Ray Computed/methods , Phantoms, Imaging , Obesity , IronABSTRACT
Objectives: To investigate the association between leisure time physical activity (LTPA) and MRI-based diastolic function and the mediating role of metabolic health. Methods: This cross-sectional analysis comprised 901 participants (46% women, mean age (SD): 56 (6) years (The Netherlands, 2008-2012)). LTPA was assessed via questionnaire, quantified in metabolic equivalent of tasks (METs)-minutes per week and participants underwent abdominal and cardiovascular MRI. Confirmatory factor analysis was used to construct the metabolic load factor. Piecewise structural equation model with adjustments for confounders was used to determine associations between LTPA and diastolic function and the mediating effect of metabolic load. Results: Significant differences in mitral early/late peak filling rate (E/A) ratio per SD of LTPA (men=1999, women=1870 MET-min/week) of 0.18, (95% CI= 0.03 to 0.33, p=0.021) were observed in men, but not in women: -0.01 (-0.01 to 0.34, p=0.058). Difference in deceleration time of mitral early filling (E-DT) was 0.13 (0.01 to 0.24, p=0.030) in men and 0.17 (0.05 to 0.28, p=0.005) in women. Metabolic load, including MRI-based visceral and subcutaneous adipose tissue, fasting glucose, high-density lipoprotein cholesterol and triglycerides, mediated these associations as follows: E/A-ratio of 0.030 (0.000 to 0.067, 19% mediated, p=0.047) in men but not in women: 0.058 (0.027 to 0.089, p<0.001) and E-DT not in men 0.004 (-0.012 to 0.021, p=0.602) but did in women 0.044 (0.013 to 0.057, 27% mediated, p=0.006). Conclusions: A larger amount of LTPA was associated with improved diastolic function where confirmatory factor analysis-based metabolic load partly mediated this effect. Future studies should assess whether improving indicators of metabolic load alongside LTPA will benefit healthy diastolic function even more.
ABSTRACT
Predicting who will benefit from treatment with immune checkpoint inhibition (ICI) in patients with advanced melanoma is challenging. We developed a multivariable prediction model for response to ICI, using routinely available clinical data including primary melanoma characteristics. We used a population-based cohort of 3525 patients with advanced cutaneous melanoma treated with anti-PD-1-based therapy. Our prediction model for predicting response within 6 months after ICI initiation was internally validated with bootstrap resampling. Performance evaluation included calibration, discrimination and internal-external cross-validation. Included patients received anti-PD-1 monotherapy (n = 2366) or ipilimumab plus nivolumab (n = 1159) in any treatment line. The model included serum lactate dehydrogenase, World Health Organization performance score, type and line of ICI, disease stage and time to first distant recurrence-all at start of ICI-, and location and type of primary melanoma, the presence of satellites and/or in-transit metastases at primary diagnosis and sex. The over-optimism adjusted area under the receiver operating characteristic was 0.66 (95% CI: 0.64-0.66). The range of predicted response probabilities was 7%-81%. Based on these probabilities, patients were categorized into quartiles. Compared to the lowest response quartile, patients in the highest quartile had a significantly longer median progression-free survival (20.0 vs 2.8 months; P < .001) and median overall survival (62.0 vs 8.0 months; P < .001). Our prediction model, based on routinely available clinical variables and primary melanoma characteristics, predicts response to ICI in patients with advanced melanoma and discriminates well between treated patients with a very good and very poor prognosis.
Subject(s)
Melanoma , Skin Neoplasms , Humans , Melanoma/pathology , Immune Checkpoint Inhibitors/therapeutic use , Skin Neoplasms/pathology , Ipilimumab/therapeutic use , Nivolumab/therapeutic use , Retrospective StudiesABSTRACT
OBJECTIVES: Cardiac motion artifacts hinder the assessment of coronary arteries in coronary computed tomography angiography (CCTA). We investigated the impact of motion compensation reconstruction (MCR) on motion artifacts in CCTA at various heart rates (HR) using a dynamic phantom. MATERIALS AND METHODS: An artificial hollow coronary artery (5-mm diameter lumen) filled with iodinated contrast agent (400 HU at 120 kVp), positioned centrally in an anthropomorphic chest phantom, was scanned using a dual-layer spectral detector CT. The artery was translated at constant horizontal velocities (0-80 mm/s, increment of 10 mm/s). For each velocity, five CCTA scans were repeated using a clinical protocol. Motion artifacts were quantified using the in-plane motion area. Regression analysis was performed to calculate the reduction in motion artifacts provided by MCR, by division of the slopes of non-MCR and MCR fitted lines. RESULTS: Reference mean (95% confidence interval) motion artifact area was 24.9 mm2 (23.8, 26.0). Without MCR, motion artifact areas for velocities exceeding 20 mm/s were significantly larger (up to 57.2 mm2 (40.1, 74.2)) than the reference. With MCR, no significant differences compared to the reference were shown for all velocities, except for 70 mm/s (29.0 mm2 (27.0, 31.0)). The slopes of the fitted data were 0.44 and 0.04 for standard and MCR reconstructions, respectively, resulting in an 11-time motion artifact reduction. CONCLUSION: MCR may improve CCTA assessment in patients by reducing coronary artery motion artifacts, especially in those with elevated HR who cannot receive beta blockers or do not attain the targeted HR. CLINICAL RELEVANCE STATEMENT: This vendor-specific motion compensation reconstruction may improve coronary computed tomography angiography assessment in patients by reduction of coronary artery motion artifacts, especially in those with elevated various heart rates (HR) who cannot receive beta blockers or do not attain the targeted HR. KEY POINTS: ⢠Motion artifacts are known to hinder the assessment of coronary arteries on coronary CT angiography (CCTA), leading to more non-diagnostic scans. ⢠This dynamic phantom study shows that motion compensation reconstruction (MCR) reduces motion artifacts at various velocities, which may help to decrease the number of non-diagnostic scans. ⢠MCR in this study showed to reduce motion artifacts 11-fold.