Cardiac Cine Magnetic Resonance Imaging Research Articles

DeepMesh: Mesh-Based Cardiac Motion Tracking Using Deep Learning.

3D motion estimation from cine cardiac magnetic resonance (CMR) images is important for the assessment of cardiac function and the diagnosis of cardiovascular diseases. Current state-of-the art methods focus on estimating dense pixel-/voxel-wise motion fields in image space, which ignores the fact that motion estimation is only relevant and useful within the anatomical objects of interest, e.g., the heart. In this work, we model the heart as a 3D mesh consisting of epi- and endocardial surfaces. We propose a novel learning framework, DeepMesh, which propagates a template heart mesh to a subject space and estimates the 3D motion of the heart mesh from CMR images for individual subjects. In DeepMesh, the heart mesh of the end-diastolic frame of an individual subject is first reconstructed from the template mesh. Mesh-based 3D motion fields with respect to the end-diastolic frame are then estimated from 2D short- and long-axis CMR images. By developing a differentiable mesh-to-image rasterizer, DeepMesh is able to leverage 2D shape information from multiple anatomical views for 3D mesh reconstruction and mesh motion estimation. The proposed method estimates vertex-wise displacement and thus maintains vertex correspondences between time frames, which is important for the quantitative assessment of cardiac function across different subjects and populations. We evaluate DeepMesh on CMR images acquired from the UK Biobank. We focus on 3D motion estimation of the left ventricle in this work. Experimental results show that the proposed method quantitatively and qualitatively outperforms other image-based and mesh-based cardiac motion tracking methods.

STADNet: Spatial-Temporal Attention-Guided Dual-Path Network for cardiac cine MRI super-resolution

Cardiac cine magnetic resonance imaging (MRI) is a commonly used clinical tool for evaluating cardiac function and morphology. However, its diagnostic accuracy may be compromised by the low spatial resolution. Current methods for cine MRI super-resolution reconstruction still have limitations. They typically rely on 3D convolutional neural networks or recurrent neural networks, which may not effectively capture long-range or non-local features due to their limited receptive fields. Optical flow estimators are also commonly used to align neighboring frames, which may cause information loss and inaccurate motion estimation. Additionally, pre-warping strategies may involve interpolation, leading to potential loss of texture details and complicated anatomical structures. To overcome these challenges, we propose a novel Spatial-Temporal Attention-Guided Dual-Path Network (STADNet) for cardiac cine MRI super-resolution. We utilize transformers to model long-range dependencies in cardiac cine MR images and design a cross-frame attention module in the location-aware spatial path, which enhances the spatial details of the current frame by using complementary information from neighboring frames. We also introduce a recurrent flow-enhanced attention module in the motion-aware temporal path that exploits the correlation between cine MRI frames and extracts the motion information of the heart. Experimental results demonstrate that STADNet outperforms SOTA approaches and has significant potential for clinical practice.

A Deep Learning-Based Integrated Framework for Quality-Aware Undersampled Cine Cardiac MRI Reconstruction and Analysis.

Cine cardiac magnetic resonance (CMR) imaging is considered the gold standard for cardiac function evaluation. However, cine CMR acquisition is inherently slow and in recent decades considerable effort has been put into accelerating scan times without compromising image quality or the accuracy of derived results. In this article, we present a fully-automated, quality-controlled integrated framework for reconstruction, segmentation and downstream analysis of undersampled cine CMR data. The framework produces high quality reconstructions and segmentations, leading to undersampling factors that are optimised on a scan-by-scan basis. This results in reduced scan times and automated analysis, enabling robust and accurate estimation of functional biomarkers. To demonstrate the feasibility of the proposed approach, we perform simulations of radial k-space acquisitions using in-vivo cine CMR data from 270 subjects from the UK Biobank (with synthetic phase) and in-vivo cine CMR data from 16 healthy subjects (with real phase). The results demonstrate that the optimal undersampling factor varies for different subjects by approximately 1 to 2 seconds per slice. We show that our method can produce quality-controlled images in a mean scan time reduced from 12 to 4 seconds per slice, and that image quality is sufficient to allow clinically relevant parameters to be automatically estimated to lie within 5% mean absolute difference.

Cardiovascular magnetic resonance imaging-derived intraventricular pressure gradients in ST-segment elevation myocardial infarction: a long-term follow-up study

Abstract Aims Recently, novel post-processing tools have become available that measure intraventricular pressure gradients (IVPGs) on routinely obtained long-axis cine cardiac magnetic resonance (CMR) images. IVPGs provide a comprehensive overview of both systolic and diastolic left ventricular (LV) functions. Whether IVPGs are associated with clinical outcome after ST-elevation myocardial infarction (STEMI) is currently unknown. Here, we investigated the association between CMR-derived LV-IVPGs and major adverse cardiovascular events (MACE) in a large reperfused STEMI cohort with long-term outcome. Methods and results In this prospectively enrolled multi-centre cohort study, 307 patients underwent CMR within 14 days after the first STEMI. LV-IVPGs (from apex-to-base) were estimated on the long-axis cine images. During a median follow-up of 9.7 (5.9–12.5) years, MACE (i.e. composite of cardiovascular death and de novo heart failure hospitalisation) occurred in 49 patients (16.0%). These patients had larger infarcts, more often microvascular injury, and impaired LV-IVPGs. In univariable Cox regression, overall LV-IVPG was significantly associated with MACE and remained significantly associated after adjustment for common clinical risk factors (hazard ratio (HR) 0.873, 95% confidence interval (CI) 0.794–0.961, P = 0.005) and myocardial injury parameters (HR 0.906, 95% CI 0.825–0.995, P = 0.038). However, adjusted for LV ejection fraction and LV global longitudinal strain (GLS), overall LV-IVPG does not provide additional prognostic information (HR 0.959, 95% CI 0.866–1.063, P = 0.426). Conclusion Early after STEMI, CMR-derived LV-IVPGs are univariably associated with MACE and this association remains significant after adjustment for common clinical risk factors and measures of infarct severity. However, LV-IVPGs do not add prognostic value to LV ejection fraction and LV GLS.

Kiosk 6R-FA-05 - Cardiac Magnetic Resonance Cine Images Derived-radiomics for the Prediction of Event FBee Survival in Patients with St-segment Elevation Myocardial Infarction

Kiosk 6R-FA-05 - Cardiac Magnetic Resonance Cine Images Derived-radiomics for the Prediction of Event FBee Survival in Patients with St-segment Elevation Myocardial Infarction

Kiosk 8R-FB-08 - Automated Cloud-deployable Pipeline for 3D Cardiac Shape Phenomics FBom 2D Cine Cardiac Magnetic Resonance Imaging

Kiosk 8R-FB-08 - Automated Cloud-deployable Pipeline for 3D Cardiac Shape Phenomics FBom 2D Cine Cardiac Magnetic Resonance Imaging

Optical Flow-Guided Cine MRI Segmentation With Learned Corrections.

In cardiac cine magnetic resonance imaging (MRI), the heart is repeatedly imaged at numerous time points during the cardiac cycle. Frequently, the temporal evolution of a certain region of interest such as the ventricles or the atria is highly relevant for clinical diagnosis. In this paper, we devise a novel approach that allows for an automatized propagation of an arbitrary region of interest (ROI) along the cardiac cycle from respective annotated ROIs provided by medical experts at two different points in time, most frequently at the end-systolic (ES) and the end-diastolic (ED) cardiac phases. At its core, a 3D TV- L1 -based optical flow algorithm computes the apparent motion of consecutive MRI images in forward and backward directions. Subsequently, the given terminal annotated masks are propagated by this bidirectional optical flow in 3D, which results, however, in improper initial estimates of the segmentation masks due to numerical inaccuracies. These initially propagated segmentation masks are then refined by a 3D U-Net-based convolutional neural network (CNN), which was trained to enforce consistency with the forward and backward warped masks using a novel loss function. Moreover, a penalization term in the loss function controls large deviations from the initial segmentation masks. This method is benchmarked both on a new dataset with annotated single ventricles containing patients with severe heart diseases and on a publicly available dataset with different annotated ROIs. We emphasize that our novel loss function enables fine-tuning the CNN on a single patient, thereby yielding state-of-the-art results along the complete cardiac cycle.

AI approach to biventricular function assessment in cine-MRI: an ultra-small training dataset and multivendor study

Objective. It was a great challenge to train an excellent and generalized model on an ultra-small data set composed of multi-orientation cardiac cine magnetic resonance imaging (MRI) images. We try to develop a 3D deep learning method based on an ultra-small training data set from muti-orientation cine MRI images and assess its performance of automated biventricular structure segmentation and function assessment in multivendor. Approach. We completed the training and testing of our deep learning networks using only heart datasets of 150 cases (90 cases for training and 60 cases for testing). This datasets were obtained from three different MRI vendors and each subject included two phases of the cardiac cycle and three cine sequences. A 3D deep learning algorithm combining Transformers and U-Net was trained. The performance of the segmentation was evaluated using the Dice metric and Hausdorff distance (HD). Based on this, the manual and automatic results of cardiac function parameters were compared with Pearson correlation, intraclass correlation coefficient (ICC) and Bland–Altman analysis in multivendor. Main results. The results show that the average Dice of 0.92, 0.92, 0.94 and HD95 of 2.50, 1.36, 1.37 for three sequences. The automatic and manual results of seven parameters were excellently correlated with the lowest r2 value of 0.824 and the highest of 0.983. The ICC (0.908–0.989, P < 0.001) showed that the results were highly consistent. Bland–Altman with a 95% limit of agreement showed there was no significant difference except for the difference in RVESV (P = 0.005) and LVM (P < 0.001). Significance. The model had high accuracy in segmentation and excellent correlation and consistency in function assessment. It provides a fast and effective method for studying cardiac MRI and heart disease.

Developing and evaluating a chronic ischemic cardiomyopathy in swine model by rest and stress CMR.

A large animal model of chronic coronary artery disease (CAD) is crucial for the understanding the underlying pathophysiological processes of chronic CAD and consequences for cardiac structure and function. The goal of this study was to develop a chronic model of CAD in a swine model and to evaluate the changes of myocardial structure, myocardial motility, and myocardial viability during coronary stenosis. A total of 30 swine (including 24 experimental animals and 6 controls) were enrolled. The chronic ischemia model was constructed by using Ameroid constrictor in experimental group. The 24 experimental animals were further divided into 4 groups (6 animals in each group) and were sacrificed at 1, 2, 3 and 4 weeks after operation for pathological examination, respectively. Cardiac magnetic resonance (CMR) was performed preoperatively and weekly postoperatively until sacrificed both in experimental and control group. CMR cine images, rest/adenosine triphosphate (ATP) stress myocardial contrast perfusion and LGE were performed and analyzed. The rest wall thickening (WT) score was calculated from rest cine images. The MPRI (myocardial perfusion reserve index) and MPR (myocardial perfusion reserve) were calculated based on rest and stress perfusion images. Pathology staining including triphenyltetrazolium chloride, HE and picrosirus red staining were performed after swine were sacrificed and collagen volume fraction (CVF) was calculated. The time to formation of ischemic, hibernating, and infarcted myocardium was recorded. In experimental group, from 1w to 4w after surgery, the rest WT score decreased gradually from 35.2 ± 2.0%, 32.0 ± 2.9% to 30.5 ± 3.0% and finally 29.06 ± 1.78%, p < 0.001. Left ventricular ejection fraction was gradually impaired after modeling (58.9 ± 12.6%, 56.3 ± 10.1%, 55.3 ± 9.0%, 53.8 ± 9.9%, respectively). And the MPR and MPRI also decreased stepwise with extent of surgery time (MPRI dropped from 2.1 ± 0.4, 2.0 ± 0.2 to 1.8 ± 0.3 and finally 1.7 ± 0.1, p = 0.004; MPR dropped from 2.3 ± 0.4, 2.1 ± 0.2 to 1.9 ± 0.4 and finally 1.8 ± 0.1, p < 0.001). Stronger associations between MPR, MPRI and CVF were paralleled lower wall thickening scores in fibrosis-affected areas. The ischemic myocardium was first appeared in the first week after surgery (involving ten segments), hibernated myocardium was first appeared in the second week after surgery (involving seventeen segments). LGE was first appeared in eight swine in the third weeks after surgery (16 segments). At 4w after surgery, average 9.6g scar tissue was found among 6 swine. At the same time, histological analysis established the presence of fibrosis and ongoing apoptosis in the infarcted area. In conclusion, our study provided valuable insights into the pathophysiological processes of chronic CAD and its consequences for cardiac structure and function in a large animal model through combining myocardial motion and stress perfusion.

Abstract 14417: Regional 4D Cardiac Magnetic Resonance Strain Associates With Late Gadolinium Enhancement in Duchenne Muscular Dystrophy Associated Cardiomyopathy

Introduction: Cardiomyopathy (CM) is the leading cause of death in Duchenne muscular dystrophy (DMD). Onset and clinical progression are varied. Cardiac magnetic resonance (CMR) late gadolinium enhancement (LGE) can detect fibrofatty replacement prior to decreased left ventricular ejection fraction (LVEF) but requires administration of gadolinium contrast agents. Non-contrast methods to characterize CM onset and progression could decrease scan time and exposure to gadolinium. Using a novel 4D (3D+time) regional strain analysis method, we hypothesized that the regional magnitude of surface area strain (E a ) would be significantly decreased in DMD patients with LGE and that E a would correlate with decreased LVEF . Methods: Short-axis cine CMR images for 43 pediatric DMD patients (13.6 ± 4.2 years) were compiled into 4D sequences for regional strain analysis (Fig. 1A). For all 4D sequences, epi- and endocardial borders were segmented across one cardiac cycle for kinematic analysis using a custom-built MATLAB graphical user interface (MathWorks, Natick, MA). Regional E a was measured by determining the relative change in the localized 4D endocardial surface mesh compared to end-diastole. Unpaired t-test and Spearman’s rho were used to determine statistical significance and correlation. Results: DMD patients had a range of CM severity: 13 with reduced LVEF of &lt;55% (30% of total) and 28 with LGE (65% of total). Peak E a was significantly different between patients with and without LGE in the basal inferior ( p =0.022), basal inferolateral ( p &lt;0.001), basal anterolateral ( p =0.002), mid inferolateral ( p =0.004), mid anterolateral ( p =0.018), and apical anterior ( p &lt;0.032), regions (Fig.1B). Basal E a also correlated significantly with LVEF ( rho =-0.45, p&lt; 0.001; Fig. 1C). Conclusion: Regional E a from 4D CMR in DMD was reduced in patients with LGE compared to those without LGE particularly in the basolateral regions. Peak basal E a correlated with a reduction in LVEF. LVEF.

Automatic classification of heart failure based on Cine-CMR images.

Heart failure (HF) is a serious and complex syndrome with a high mortality rate. In clinical diagnosis, the correct classification of HF is helpful. In our previous work, we proposed a self-supervised learning framework of HF classification (SSLHF) on cine cardiac magnetic resonance images (Cine-CMR). However, this method lacks the integration of three dimensions of spatial information and temporal information. Thus, this study aims at proposing an automatic 4D HF classification algorithm. To construct a 4D classification model, we proposed an extensional framework called 4D-SSLHF. It mainly consists of self-supervised image restoration and HF classification. The image restoration proxy task utilizes three image transformation methods to enhance the exploration of spatial and temporal information in the Cine-CMR. In the classification task, we proposed a Siamese Conv-LSTM network by combining the Siamese network and bi-directional Conv-LSTM to integrate the features of the four dimensions simultaneously. Experimental results on 184 patients from Shanghai Chest Hospital achieved an AUC of 0.8794 and an ACC of 0.8402 in the five-fold cross-validation. Compared with our previous work, the improvements in AUC and ACC were 2.89 % and 1.94 %, respectively. In this study, we proposed a novel self-supervised learning framework named 4D-SSLHF for HF classification based on Cine-CMR. The proposed 4D-SSLHF can mine 3D spatial information and temporal information in Cine-CMR images well and accurately classify different categories of HF. The good classification results show our method's potential to assist physicians in choosing personalized treatment.

Integrin α4 Positive Subpopulation in Adipose Derived Stem Cells Effectively Reduces Infarct Size through Enhanced Engraftment into Myocardial Infarction.

The efficacy of adipose-derived stem cells (ASCs) on myocardial infarction is limited due to poor survival and engraftment. Integrin-mediated cell adhesion is a prerequisite for its survival and homing. ASCs expressed insufficient integrin α4, limiting their homing capacity. This study aims to characterize integrin α4＋ ASC subpopulation and investigate their therapeutic efficacy in myocardial infarction. We used fluorescence-activated cell sorting to harvest integrin α4＋ ASCs subpopulation, which were characterized in vitro and transplanted into myocardial infarction model. Positron emission tomography imaging were performed to measure infarction size. Cardiac cine magnetic resonance imaging was used to evaluate heart contractile function. Compared with the unfractionated ASCs, integrin α4＋ ASCs subpopulation secreted a higher level of angiogenic growth factors, migrated more rapidly, and exhibited a stronger anti-apoptotic capacity. Vascular cell adhesion molecule-1 was obviously up-regulated at 3 days after myocardial infarction, which interacted with integrin α4 receptor on the surface of ASCs to enhance the survival and adhesion. Thus, we implanted unfractionated ASCs or integrin α4＋ ASCs subpopulation into the 3-day infarcted myocardium. Integrin α4＋ ASCs subpopulation exhibited more robust engraftment into the infarcted myocardium. Integrin α4＋ ASCs subpopulation more effectively decreased infarct size and strengthen cardiac function recovery than did the unfractionated ASCs. Integrin α4＋ ASCs subpopulation is superior to unfractionated ASCs in ameliorating ischemic myocardial damage in animal model. Mechanistically, their more robust engraftment into the infarct area, higher migratory capacity and their increased release of paracrine factors contribute to enhanced tissue repair.

Diagnostic accuracy for CZT gamma camera compared to conventional gamma camera technique with myocardial perfusion single-photon emission computed tomography: Assessment of myocardial infarction and function

BackgroundThe solid-state cadmium-zinc-telluride (CZT) gamma camera for myocardial perfusion single-photon emission computed tomography (MPS) has theoretical advantages compared to the conventional gamma camera technique. This includes more sensitive detectors and better energy resolution. We aimed to explore the diagnostic performance of gated MPS with a CZT gamma camera compared to a conventional gamma camera for detection of myocardial infarct (MI) and assessment of left ventricular (LV) volumes and ejection fraction (LVEF), using cardiac magnetic resonance (CMR) as the reference method. MethodsSeventy-three patients (26% female) with known or suspected chronic coronary syndrome were examined with gated MPS using both a CZT gamma camera and a conventional gamma camera as well as with CMR. Presence and extent of MI on MPS and late gadolinium enhancement (LGE) CMR was evaluated. For LV volumes, LVEF and LV mass, gated MPS images and cine CMR images were evaluated. ResultsMI was found in 42 patients on CMR. The overall sensitivity, specificity, positive and negative predictive values for the CZT and the conventional gamma camera were the same (67%, 100%, 100% and 69%). For infarct size > 3% on CMR, the sensitivity was 82% for the CZT and 73% for the conventional gamma camera, respectively. LV volumes were significantly underestimated by MPS compared to CMR (P ≤ .002 for all measures). The underestimation was slightly less pronounced for the CZT compared to the conventional gamma camera (2-10 mL, P ≤ .03 for all measures). For LVEF, however, accuracy was high for both gamma cameras. ConclusionDifferences between a CZT and a conventional gamma camera for detection of MI and assessment of LV volumes and LVEF are small and do not appear to be clinically significant.

Role of TRPC3 in Right Ventricular Dilatation under Chronic Intermittent Hypoxia in 129/SvEv Mice

Patients with obstructive sleep apnea (OSA) exhibit a high prevalence of pulmonary hypertension and right ventricular (RV) hypertrophy. However, the exact molecule responsible for the pathogenesis remains unknown. Given the resistance to RV dilation observed in transient receptor potential canonical 3(Trpc3)−/− mice during a pulmonary hypertension model induced by phenylephrine (PE), we hypothesized that TRPC3 also plays a role in chronic intermittent hypoxia (CIH) conditions, which lead to RV dilation and dysfunction. To test this, we established an OSA mouse model using 8- to 12-week-old 129/SvEv wild-type and Trpc3−/− mice in a customized breeding chamber that simulated sleep and oxygen cycles. Functional parameters of the RV were evaluated through analysis of cardiac cine magnetic resonance images, while histopathological examinations were conducted on cardiomyocytes and pulmonary vessels. Following exposure to 4 weeks of CIH, Trpc3−/− mice exhibited significant RV dysfunction, characterized by decreased ejection fraction, increased end-diastole RV wall thickness, and elevated expression of pathological cardiac markers. In addition, reactive oxygen species (ROS) signaling and the endothelin system were markedly increased solely in the hearts of CIH-exposed Trpc3−/− mice. Notably, no significant differences in pulmonary vessel thickness or the endothelin system were observed in the lungs of wild-type (WT) and Trpc3−/− mice subjected to 4 weeks of CIH. In conclusion, our findings suggest that TRPC3 serves as a regulator of RV resistance in response to pressure from the pulmonary vasculature, as evidenced by the high susceptibility to RV dilation in Trpc3−/− mice without notable changes in pulmonary vasculature under CIH conditions.

Adaptive feature aggregation based multi-task learning for uncertainty-guided semi-supervised medical image segmentation

Automatic segmentation of medical images is a necessary prerequisite for diagnosing related diseases. Magnetic resonance imaging (MRI) is a widely used non-invasive method in clinical practice, but obtaining reliable labeled training data for deep learning models is time-consuming and challenging. In this study, we propose an adaptive feature aggregation-based multi-task learning for uncertainty-guided semi-supervised image segmentation, dubbed AFAM-Net. The framework leverages a reconstruction task to capture anatomical information from raw images and assist the segmentation task. Moreover, we propose an adaptive feature aggregation strategy that selectively transfers useful features while filtering out irrelevant information, considering the associations between the two tasks. To better utilize unlabeled data, we incorporate dual uncertainty-aware methods to improve the segmentation performance. We evaluate the proposed AFAM-Net on clinical liver, 3D heart, and Cine cardiac MRI datasets. According to the experimental results, our AFAM-Net significantly outperformed other state-of-the-art semi-supervised medical image segmentation algorithms.

A new compressed sensing cine cardiac MRI sequence with free-breathing real-time acquisition and fully automated motion-correction: A comprehensive evaluation

PurposeThe purpose of this study was to compare a new free-breathing compressed sensing cine (FB-CS) cardiac magnetic resonance imaging (CMR) to the standard reference multi-breath-hold segmented cine (BH-SEG) CMR in an unselected population. Materials and methodsFrom January to April 2021, 52 consecutive adult patients who underwent both conventional BH-SEG CMR and new FB-CS CMR with fully automated respiratory motion correction were retrospectively enrolled. There were 29 men and 23 women with a mean age of 57.7 ± 18.9 (standard deviation [SD]) years (age range: 19.0–90.0 years) and a mean cardiac rate of 74.6 ± 17.9 (SD) bpm. For each patient, short-axis stacks were acquired with similar parameters providing a spatial resolution of 1.8 × 1.8 × 8.0 mm3 and 25 cardiac frames. Acquisition and reconstruction times, image quality (Likert scale from 1 to 4), left and right ventricular volumes and ejection fractions, left ventricular mass, and global circumferential strain were assessed for each sequence. ResultsFB-CS CMR acquisition time was significantly shorter (123.8 ± 28.4 [SD] s vs. 267.2 ± 39.3 [SD] s for BH-SEG CMR; P < 0.0001) at the penalty of a longer reconstruction time (271.4 ± 68.7 [SD] s vs. 9.9 ± 2.1 [SD] s for BH-SEG CMR; P < 0.0001). In patients without arrhythmia or dyspnea, FB-CS CMR provided subjective image quality that was not different from that of BH-SEG CMR (P = 0.13). FB-CS CMR improved image quality in patients with arrhythmia (n = 18; P = 0.002) or dyspnea (n = 7; P = 0.02), and the edge sharpness was improved at end-systole and end-diastole (P = 0.0001). No differences were observed between the two techniques in ventricular volumes and ejection fractions, left ventricular mass or global circumferential strain in patients in sinus rhythm or with cardiac arrhythmia. ConclusionThis new FB-CS CMR addresses respiratory motion and arrhythmia-related artifacts without compromising the reliability of ventricular functional assessment.

Automatic right ventricular segmentation for cine cardiac magnetic resonance images based on a new deep atlas network.

The high morbidity and mortality of heart disease present a significant threat to human health. The development of methods for the quick and accurate diagnosis of heart diseases, enabling their effective treatment, has become a key issue of concern. Right ventricular (RV) segmentation from cine cardiac magnetic resonance (CMR) images plays a significant role in evaluating cardiac function for clinical diagnosis and prognosis. However, due to the complex structure of the RV, traditional segmentation methods are ineffective for RV segmentation. In this paper, we propose a new deep atlas network to improve the learning efficiency and segmentation accuracy of a deep learning network by integrating multi-atlas. First, a dense multi-scale U-net (DMU-net) is presented to acquire transformation parameters from atlas images to target images. The transformation parameters map the atlas image labels to the target image labels. Second, using a spatial transformation layer, the atlas images are deformed based on these parameters. Finally, the network is optimized by backpropagation with two loss functions where the mean squared error function (MSE) is used to measure the similarity of the input images and transformed images. Further, the Dice metric (DM) is used to quantify the overlap between the predicted contours and the ground truth. In our experiments, 15 datasets are used in testing, and 20 cine CMR images are selected as atlas. The mean values and standard deviations for the DM and Hausdorff distance are 0.871 and 4.67mm, 0.104 and 2.528mm, respectively. The correlation coefficients of endo-diastolic volume, endo-systolic volume, ejection fraction, and stroke volume are 0.984, 0.926, 0.980, and 0.991, respectively, and the mean differences between all of the mentioned parameters are 3.2, -1.7, 0.02, and 4.9, respectively. Most of these differences are within the allowable range of 95%, indicating that the results are acceptable and show good consistency. The segmentation results obtained in this method are compared with those obtained by other methods that provide satisfactory performance. The other methods provide better segmentation effects at the base, but either no segmentation or the wrong segmentation at the top, which demonstrate that the deep atlas network can improve top-area segmentation accuracy. Our results indicate that the proposed method can achieve better segmentation results than the previous methods, with both high relevance and consistency, and has the potential for clinical application.

Deep learning-based prognostic model using non-enhanced cardiac cine MRI for outcome prediction in patients with heart failure.

To evaluate the performance of a deep learning-based multi-source model for survival prediction and risk stratification in patients with heart failure. Patients with heart failure with reduced ejection fraction (HFrEF) who underwent cardiac magnetic resonance between January 2015 and April 2020 were retrospectively included in this study. Baseline electronic health record data, including clinical demographic information, laboratory data, and electrocardiographic information, were collected. Short-axis non-contrast cine images of the whole heart were acquired to estimate the cardiac function parameters and the motion features of the left ventricle. Model accuracy was evaluated using the Harrell's concordance index. All patients were followed up for major adverse cardiac events (MACEs), and survival prediction was assessed using Kaplan-Meier curves. A total of 329 patients were evaluated (age 54 ± 14years; men, 254) in this study. During a median follow-up period of 1041days, 62 patients experienced MACEs and their median survival time was 495days. When compared with conventional Cox hazard prediction models, deep learning models showed better survival prediction performance. Multi-data denoising autoencoder (DAE) model reached the concordance index of 0.8546 (95% CI: 0.7902-0.8883). Furthermore, when divided into phenogroups, the multi-data DAE model could significantly discriminate between the survival outcomes of the high-risk and low-risk groups compared with other models (p < 0.001). The proposed deep learning (DL) model based on non-contrast cardiac cine magnetic resonance imaging could independently predict the outcome of patients with HFrEF and showed better prediction efficiency than conventional methods. The proposed multi-source deep learning model based on cardiac magnetic resonance enables survival prediction in patients with heart failure. • A multi-source deep learning model based on non-contrast cardiovascular magnetic resonance (CMR) cine images was built to make robust survival prediction in patients with heart failure. • The ground truth definition contains electronic health record data as well as DL-based motion data, and cardiac motion information is extracted by optical flow method from non-contrast CMR cine images. • The DL-based model exhibits better prognostic value and stratification performance when compared with conventional prediction models and could aid in the risk stratification in patients with HF.

Characterization of biatrial sports adaption in a large, highly trained athletes cohort using feature tracking cardiac magnetic resonance

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Project no. TKP2021-NKTA-46 has been implemented with the support provided by the Ministry of Innovation and Technology of Hungary from the National Research, Development and Innovation Fund, financed under the TKP2021-NKTA funding scheme. This project was also supported by a grant from the National Research, Development, and Innovation Office (NKFIH) of Hungary (K135076 to BM). Introduction Feature tracking strain analysis using cardiac magnetic resonance (CMR) imaging enables the appreciation of complex mechanics of the athlete’s heart without adding extra sequences and increasing image acquisition time. Purpose We evaluated the associations of the left (LA) and right atrial (RA) functional and strain parameters with sports activity and demographic characteristics in a large cohort of competitive athletes. We compared the athlete’s biatrial remodelling with healthy, sex-matched control subjects. Methods We performed a retrospective analysis on native CMR cine images acquired between 2009 and 2021. Standard two-chamber and four-chamber long-axis cine movies were analyzed using Medis Suite software. We excluded images with foreshortening or apparent planning flaw. Subjects were categorized into four groups based on their weekly training loads (&amp;lt;6 hours, 6-10 hours, 11-20 hours and 21&amp;lt; hours). Results We performed CMR on 510 subjects, of which 375 were highly trained athletes (median age 20 ys, IQR 16-25; 244 male). A group of 135 healthy volunteers [median (mdn) age 25 ys, IQR 23-28, 64 male] served as controls. Left atrial global longitudinal strain (GLS) was lower (p=.002) in athletes (mdn 41%) than in controls (mdn 43%), whereas RA GLS showed a more pronounced decrease (mdn 32% vs 43%, p&amp;lt;.0001). Between male athletes and controls, LA global circumferential strain [(GCS), mdn 40% vs 47%, p=.0001], LA conduit (mdn 26% vs 28%, p=.025), RA GLS (mdn 31% vs 38%, p&amp;lt;.0001), RA GCS (mdn 17% vs 20%, p=.024), RA booster (mdn 10% vs 9%, p&amp;lt;.001) and RA conduit (mdn 21% vs 24%, p=.001) strain values showed a difference. Between female athletes and controls, RA GLS (mdn 34% vs 44%, p&amp;lt;.0001), RA GCS (mdn 19% vs 29%, p&amp;lt;.001), RA booster (mdn 9% vs 13%, p=.006), and RA conduit (mdn 25% vs 32%, p&amp;lt;.001) were significantly different, while LA strain values were comparable. We found a significant difference between male and female athletes in terms of LA GLS (mdn 39%, vs 43%, p=.0001), LA GCS (mdn 40% vs 42%, p=.01) and LA conduit strain (mdn 26% vs 29%, p&amp;lt;.001). Regarding the right atrium, GLS (mdn 31% vs 34%, p=.047) and conduit strain (mdn 21% vs 25%, p=.004) were significantly lower in male athletes. LA GLS only differed between the lowest and highest training volumes (p=0.048). LA GCS was higher in the least trained group (p&amp;lt;.001) than in the other three categories. Interestingly, the LA end-diastolic volume index was less (r=0.13, p=0.008) correlated to weekly training volume than the left ventricular end-diastolic volume index (r=0.51, p&amp;lt;.0001). Conclusion The atrial adaptation of the athlete’s heart shows distinct mechanic properties, which are further stratified with exercise level and sex. The RA seems to be more prone to increasing exercise load than the LA. Sophisticated atrial function characterization by strain analysis might provide novel insight into healthy sports adaptation.

CMR-derived left ventricular intraventricular pressure gradients identify different patterns associated with prognosis in dilated cardiomyopathy.

Left ventricular (LV) blood flow is determined by intraventricular pressure gradients (IVPG). Changes in blood flow initiate remodelling and precede functional decline. Novel cardiac magnetic resonance (CMR) post-processing LV-IVPG analysis might provide a sensitive marker of LV function in dilated cardiomyopathy (DCM). Therefore, the aim of our study was to evaluate LV-IVPG patterns and their prognostic value in DCM. LV-IVPGs between apex and base were measured on standard CMR cine images in DCM patients (n = 447) from the Maastricht Cardiomyopathy registry. Major adverse cardiovascular events, including heart failure hospitalisations, life-threatening arrhythmias, and sudden/cardiac death, occurred in 66 DCM patients (15%). A temporary LV-IVPG reversal during systolic-diastolic transition, leading to a prolonged transition period or slower filling, was present in 168 patients (38%). In 14%, this led to a reversal of blood flow, which predicted outcome corrected for univariable predictors [hazard ratio (HR) = 2.57, 95% confidence interval (1.01-6.51), P = 0.047]. In patients without pressure reversal (n = 279), impaired overall LV-IVPG [HR = 0.91 (0.83-0.99), P = 0.033], systolic ejection force [HR = 0.91 (0.86-0.96), P < 0.001], and E-wave decelerative force [HR = 0.83 (0.73-0.94), P = 0.003] predicted outcome, independent of known predictors (age, sex, New York Heart Association class ≥ 3, LV ejection fraction, late gadolinium enhancement, LV-longitudinal strain, left atrium (LA) volume-index, and LA-conduit strain). Pressure reversal during systolic-diastolic transition was observed in one-third of DCM patients, and reversal of blood flow direction predicted worse outcome. In the absence of pressure reversal, lower systolic ejection force, E-wave decelerative force (end of passive LV filling), and overall LV-IVPG are powerful predictors of outcome, independent of clinical and imaging parameters.