Predicting Late Gadolinium Enhancement of Acute Myocardial Infarction in Contrast-Free Cardiac Cine MRI Using Deep Generative Learning.

BACKGROUND: Late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) is a standard technique for diagnosing myocardial infarction (MI), which, however, poses risks due to gadolinium contrast usage. Techniques enabling MI assessment based on contrast-free CMR are desirable to overcome the limitations associated with contrast enhancement. METHODS: We introduce a novel deep generative learning method, termed cine-generated enhancement (CGE), which transforms standard contrast-free cine CMR into LGE-equivalent images for MI assessment. CGE features with multislice spatiotemporal feature extractor, enhancement contrast modulation, and sophisticated loss function. Data from 430 patients with acute MI from 3 centers were collected. After image quality control, 1525 pairs (289 patients) of center I were used for training, and 293 slices (52 patients) of the same center were reserved for internal testing. The 40 patients (401 slices) of the other 2 centers were used for external testing. The CGE robustness was further tested in 20 normal subjects in a public cine CMR data set. CGE images were compared with LGE for image quality assessment and MI quantification regarding scar size and transmurality. RESULTS: The CGE method produced images of superior quality to LGE in both internal and external data sets. There was a significant (P<0.001) correlation between CGE and LGE measurements of scar size (Pearson correlation, 0.79/0.80; intraclass correlation coefficient, 0.79/0.77) and transmurality (Pearson correlation, 0.76/0.64; intraclass correlation coefficient, 0.76/0.63) in internal/external data set. Considering all data sets, CGE demonstrated high sensitivity (91.27%) and specificity (95.83%) in detecting scars. Realistic enhancement images were obtained for the normal subjects in the public data set without false positive subjects. CONCLUSIONS: CGE achieved superior image quality to LGE and accurate scar delineation in patients with acute MI of both internal and external data sets. CGE can significantly simplify the CMR examination, reducing scan times and risks associated with gadolinium-based contrasts, which are crucial for acute patients.

The association between cardiac T2*BOLD and quantitative flow ratio (QFR) in the diagnosis of stenotic coronary arteries in patients with multi-vessel coronary artery disease.

BACKGROUND: T2*BOLD is based on myocardial deoxyhemoglobin content to reflect the state of myocardial oxygenation. Quantitative flow ratio is a tool for assessing coronary blood flow based on invasive coronary angiography. PURPOSE: This study aimed to evaluate the correlation between T2*BOLD and QFR in the diagnosis of stenotic coronary arteries in patients with multi-vessel coronary artery disease. METHODS: Fifty patients with MVCAD with at least 1 significant coronary artery stenosis (diameter stenosis > 50%) and 21 healthy control subjects underwent coronary angiography combined with QFR measurements and cardiovascular magnetic resonance (CMR). QFR ≤ 0.80 was considered to indicate the presence of hemodynamic obstruction. RESULTS: Totally 60 (54%) obstructive vessels had hemodynamic change. Between stenotic coronary arteries (QFR ≤ 0.8) and normal vessels, T2*BOLD showed AUCs of 0.97, 0.69, and 0.91 for left anterior descending (LAD), left circumflex (LCX) and right coronary (RCA) arteries and PI displayed AUCs of 0.89, 0.77 and 0.90 (all p > 0.05, except for LAD). The AUCs of T2*BOLD between stenotic coronary arteries (QFR > 0.8) and normal vessels were 0.86, 0.72, and 0.85 for LAD, LCX and RCA; while, PI showed AUCs of 0.93, 0.86, and 0.88, respectively (p > 0.05). Moreover, T2*BOLD displayed AUCs of 0.96, 0.74, and 0.91 for coronary arteries as before between coronary arteries with stenosis (QFR ≤ 0.8 and > 0.8), but the mean PI of LAD, LCX and RCA showed no significant differences between them. CONCLUSION: T2* BOLD and QFR have good correlation in diagnosing stenotic coronary arteries with hemodynamic changes in patients with stable multi-vessel CAD. T2* BOLD is superior to semi-quantitative perfusion imaging in analyzing myocardial ischemia without stress.

Fractal analysis of left ventricular trabeculae in heart failure with preserved ejection fraction patients with multivessel coronary artery disease.

OBJECTIVES: Endocardial trabeculae undergo varicose changes and hyperplasia in response to hemodynamic influences and are a variable phenotype reflecting changes in disease. Fractal analysis has been used to analyze the complexity of endocardial trabeculae in a variety of cardiomyopathies. The aim of this paper was to quantify the myocardial trabecular complexity through fractal analysis and to investigate its predictive value for the diagnosis of heart failure with preserved ejection fraction (HFpEF) in patients with multivessel coronary artery disease (CAD). METHODS: The retrospective study population consisted of 97 patients with multivessel CAD, 39 of them were diagnosed with HFpEF, while 46 healthy volunteers were recruited as controls. Fractal dimension (FD) was obtained through fractal analysis of endocardial trabeculae on LV short-axis cine images. Logistic regression analyses were used to confirm the predictors and compare different prediction models. RESULTS: Mean basal FD was significantly higher in patients with HFpEF than in patients without HFpEF or in the healthy group (median: 1.289; IQR: 0.078; p < 0.05). Mean basal FD was also a significant independent predictor in univariate and multivariate logistic regression (OR: 1.107 and 1.043, p < 0.05). Furthermore, adding FD to the prediction model improved the calibration and accuracy of the model (c-index: 0.806). CONCLUSION: The left ventricular FD obtained with fractal analysis can reflect the complexity of myocardial trabeculae and has an independent predictive value for the diagnosis of HFpEF in patients with multivessel CAD. Including FD into the diagnostic model can help improve the diagnosis. CRITICAL RELEVANCE STATEMENT: Differences show in the complexity of endocardial trabeculae in multivessel coronary artery disease patients, and obtaining fractal dimensions (FD) by fractal analysis can help identify heart failure with preserved ejection fraction (HFpEF) patients. KEY POINTS: The complexity of myocardial trabeculae differs among patients with multivessel coronary artery disease. Left ventricular fractal dimensions can reflect the complexity of the myocardial trabecular. Fractal dimensions have predictive value for the diagnosis of heart failure with preserved ejection fraction.

Coronary artery wall contrast enhancement imaging impact on disease activity assessment in IgG4-RD: a direct marker of coronary involvement.

BACKGROUND: Coronary artery wall contrast enhancement (CE) has been applied to non-invasive visualization of changes to the coronary artery wall in systemic lupus erythematosus (SLE). This study investigated the feasibility of quantifying CE to detect coronary involvement in IgG4-related disease (IgG4-RD), as well as the influence on disease activity assessment. METHODS: A total of 93 subjects (31 IgG4-RD; 29 SLE; 33 controls) were recruited in the study. Coronary artery wall imaging was performed in a 3.0 T MRI scanner. Serological markers and IgG4-RD Responder Index (IgG4-RD-RI) scores were collected for correlation analysis. RESULTS: Coronary wall CE was observed in 29 (94 %) IgG4-RD patients and 22 (76 %) SLE patients. Contrast-to-noise ratio (CNR) and total CE area were significantly higher in patient groups compared to controls (CNR: 6.1 ± 2.7 [IgG4-RD] v. 4.2 ± 2.3 [SLE] v. 1.9 ± 1.5 [control], P < 0.001; Total CE area: 3.0 [3.0-6.6] v. 1.7 [1.5-2.6] v. 0.3 [0.3-0.9], P < 0.001). In the IgG4-RD group, CNR and total CE area were correlated with the RI (CNR: r = 0.55, P = 0.002; total CE area: r = 0.39, P = 0.031). RI´ scored considering coronary involvement by CE, differed significantly from RI scored without consideration of CE (RI v. RI´: 15 ± 6 v. 16 ± 6, P < 0.001). CONCLUSIONS: Visualization and quantification of CMR coronary CE by CNR and total CE area could be utilized to detect subclinical and clinical coronary wall involvement, which is prevalent in IgG4-RD. The potential inclusion of small and medium-sized vessel involvements in the assessment of disease activity in IgG4-RD is worthy of further investigation.

Ligand-Induced Atomically Segregation-Tunable Alloy Nanoprobes for Enhanced Magnetic Resonance Imaging.

Bimetallic iron-noble metal alloy nanoparticles have emerged as promising contrast agents for magnetic resonance imaging (MRI) due to their biocompatibility and facile control over the element distribution. However, the inherent surface energy discrepancy between iron and noble metal often leads to Fe atom segregation within the nanoparticle, resulting in limited iron-water molecule interactions and, consequently, diminished relaxometric performance. In this study, we present the development of a class of ligand-induced atomically segregation-tunable alloy nanoprobes (STAN) composed of bimetallic iron-gold nanoparticles. By manipulating the oxidation state of Fe on the particle surface through varying molar ratios of oleic acid and oleylamine ligands, we successfully achieve surface Fe enrichment. Under the application of a 9 T MRI system, the optimized STAN formulation, characterized by a surface Fe content of 60.1 at %, exhibits an impressive r1 value of 2.28 mM-1·s-1, along with a low r2/r1 ratio of 6.2. This exceptional performance allows for the clear visualization of hepatic tumors as small as 0.7 mm in diameter in vivo, highlighting the immense potential of STAN as a next-generation contrast agent for highly sensitive MR imaging.

Screening and diagnosis of cardiovascular disease using artificial intelligence-enabled cardiac magnetic resonance imaging.

Cardiac magnetic resonance imaging (CMR) is the gold standard for cardiac function assessment and plays a crucial role in diagnosing cardiovascular disease (CVD). However, its widespread application has been limited by the heavy resource burden of CMR interpretation. Here, to address this challenge, we developed and validated computerized CMR interpretation for screening and diagnosis of 11 types of CVD in 9,719 patients. We propose a two-stage paradigm consisting of noninvasive cine-based CVD screening followed by cine and late gadolinium enhancement-based diagnosis. The screening and diagnostic models achieved high performance (area under the curve of 0.988 ± 0.3% and 0.991 ± 0.0%, respectively) in both internal and external datasets. Furthermore, the diagnostic model outperformed cardiologists in diagnosing pulmonary arterial hypertension, demonstrating the ability of artificial intelligence-enabled CMR to detect previously unidentified CMR features. This proof-of-concept study holds the potential to substantially advance the efficiency and scalability of CMR interpretation, thereby improving CVD screening and diagnosis.

Epicardial adipose tissue in obesity with heart failure with preserved ejection fraction: Cardiovascular magnetic resonance biomarker study.

BACKGROUND: Obesity has become a serious public health issue, significantly elevating the risk of various complications. It is a well-established contributor to Heart failure with preserved ejection fraction (HFpEF). Evaluating HFpEF in obesity is crucial. Epicardial adipose tissue (EAT) has emerged as a valuable tool for validating prognostic biomarkers and guiding treatment targets. Hence, assessing EAT is of paramount importance. Cardiovascular magnetic resonance (CMR) imaging is acknowledged as the gold standard for analyzing cardiac function and morphology. We hope to use CMR to assess EAT as a bioimaging marker to evaluate HFpEF in obese patients. AIM: To assess the diagnostic utility of CMR for evaluating heart failure with preserved ejection fraction [HFpEF; left ventricular (LV) ejection fraction ≥ 50%] by measuring the epicardial adipose tissue (EAT) volumes and EAT mass in obese patients. METHODS: Sixty-two obese patients were divided into two groups for a case-control study based on whether or not they had heart failure with HFpEF. The two groups were defined as HFpEF+ and HFpEF-. LV geometry, global systolic function, EAT volumes and EAT mass of all subjects were obtained using cine magnetic resonance sequences. RESULTS: Forty-five patients of HFpEF- group and seventeen patients of HFpEF+ group were included. LV mass index (g/m2) of HFpEF+ group was higher than HFpEF- group (P < 0.05). In HFpEF+ group, EAT volumes, EAT volume index, EAT mass, EAT mass index and the ratio of EAT/[left atrial (LA) left-right (LR) diameter] were higher compared to HFpEF- group (P < 0.05). In multivariate analysis, Higher EAT/LA LR diameter ratio was associated with higher odds ratio of HFpEF. CONCLUSION: EAT/LA LR diameter ratio is highly associated with HFpEF in obese patients. It is plausible that there may be utility in CMR for assessing obese patients for HFpEF using EAT/LA LR diameter ratio as a diagnostic biomarker. Further prospective studies, are needed to validate these proof-of-concept findings.

Toward Accurate Cardiac MRI Segmentation With Variational Autoencoder-Based Unsupervised Domain Adaptation.

Accurate myocardial segmentation is crucial in the diagnosis and treatment of myocardial infarction (MI), especially in Late Gadolinium Enhancement (LGE) cardiac magnetic resonance (CMR) images, where the infarcted myocardium exhibits a greater brightness. However, segmentation annotations for LGE images are usually not available. Although knowledge gained from CMR images of other modalities with ample annotations, such as balanced-Steady State Free Precession (bSSFP), can be transferred to the LGE images, the difference in image distribution between the two modalities (i.e., domain shift) usually results in a significant degradation in model performance. To alleviate this, an end-to-end Variational autoencoder based feature Alignment Module Combining Explicit and Implicit features (VAMCEI) is proposed. We first re-derive the Kullback-Leibler (KL) divergence between the posterior distributions of the two domains as a measure of the global distribution distance. Second, we calculate the prototype contrastive loss between the two domains, bringing closer the prototypes of the same category across domains and pushing away the prototypes of different categories within or across domains. Finally, a domain discriminator is added to the output space, which indirectly aligns the feature distribution and forces the extracted features to be more favorable for segmentation. In addition, by combining CycleGAN and VAMCEI, we propose a more refined multi-stage unsupervised domain adaptation (UDA) framework for myocardial structure segmentation. We conduct extensive experiments on the MSCMRSeg 2019, MyoPS 2020 and MM-WHS 2017 datasets. The experimental results demonstrate that our framework achieves superior performances than state-of-the-art methods.