Cardiac Cine Magnetic Resonance Imaging Research Articles

Segment Length in Cine Strain Analysis Predicts Cardiac Resynchronization Therapy Outcome Beyond Current Guidelines.

Patients with a class I recommendation for cardiac resynchronization therapy (CRT) are likely to benefit, but the effect of CRT in class II patients is more heterogeneous and additional selection parameters are needed in this group. The recently validated segment length in cine strain analysis of the septum (SLICE-ESSsep) measurement on cardiac magnetic resonance cine imaging predicts left ventricular functional recovery after CRT but its prognostic value is unknown. This study sought to evaluate the prognostic value of SLICE-ESSsep for clinical outcome after CRT. Two hundred eighteen patients with a left bundle branch block or intraventricular conduction delay and a class I or class II indication for CRT who underwent preimplantation cardiovascular magnetic resonance examination were enrolled. SLICE-ESSsep was manually measured on standard cardiovascular magnetic resonance cine imaging. The primary combined end point was all-cause mortality, left ventricular assist device, or heart transplantation. Secondary end points were (1) appropriate implantable cardioverter defibrillator therapy and (2) heart failure hospitalization. Two-thirds (65%) of patients had a positive SLICE-ESSsep ≥0.9% (ie, systolic septal stretching). During a median follow-up of 3.8 years, 66 (30%) patients reached the primary end point. Patients with positive SLICE-ESSsep were at lower risk to reach the primary end point (hazard ratio 0.36; P<0.001) and heart failure hospitalization (hazard ratio 0.41; P=0.019), but not for implantable cardioverter defibrillator therapy (hazard ratio, 0.66; P=0.272). Clinical outcome of class II patients with a positive ESSsep was similar to those of class I patients (hazard ratio, 1.38 [95% CI, 0.66-2.88]; P=0.396). Strain assessment of the septum (SLICE-ESSsep) provides a prognostic measure for clinical outcome after CRT. Detection of a positive SLICE-ESSsep in patients with a class II indication predicts improved CRT outcome similar to those with a class I indication whereas SLICE-ESSsep negative patients have poor prognosis after CRT implantation.

Utility of single-shot compressed sensing cardiac magnetic resonance cine imaging for assessment of biventricular function in free-breathing and arrhythmic pediatric patients

BackgroundThis study aimed to explore the feasibility and accuracy of single-shot compressed-sensing (CS) cardiac magnetic resonance cine technology for the assessment of biventricular function and morphology in free-breathing (FB) pediatrics, especially those with arrhythmia. MethodsSeventy consecutive pediatric participants (6.27 ± 3.8 years, range:0.5–14 years) were enrolled between August 2019 and July 2020. Single-shot CS and conventional balanced steady-state free-precession (bSSFP) cine were obtained. The total scanning time, image quality and biventricular function parameters were compared for both sequences. ResultsSingle-shot CS cine had shorter acquisition time compared with the conventional bSSFP cine (all P < 0.001). The single-shot CS cine also had fewer artifacts than conventional bSSFP cine (breath-hold (BH): 4.6 ± 0.6 vs. 4.3 ± 0.6; FB without ongoing arrhythmia: 4.5 ± 0.6 vs. 3.6 ± 0.9; FB with ongoing arrhythmia: 4.7 ± 0.5 vs. 2.6 ± 1.1; all P < 0.05). No statistical difference of left ventricular parameters and right ventricular end-systolic volume/ejection fraction were found between the single-shot CS and conventional bSSFP cine in both BH and FB without ongoing arrhythmia group. There was an excellent correlation (R2 = 0.60–0.98, all P < 0.001) and good intra-(range: R2 = 0.57–0.99, P < 0.001)/inter-observer agreements (range: R2 = 0.76–1, P < 0.001) for single-shot CS cine images in terms of biventricular function parameters. ConclusionsThe single-shot CS cine can significantly reduce the image acquisition time, offering reliable quantification of biventricular function in free breathing condition for arrhythmic patients.

Innovative procedure for measuring left ventricular ejection fraction from 18F-FDG first-pass ultra-sensitive digital PET/CT images: evaluation with an anthropomorphic heart phantom

BackgroundLeft ventricular ejection fraction (LVEF) is usually measured by cine-cardiac magnetic resonance imaging (MRI), planar and single-photon emission-computerized tomography (SPECT) equilibrium radionuclide angiocardiography (ERNA), and echocardiography. It would be clinically useful to measure LVEF from first-pass positron-emission tomography/computed tomography (PET/CT) radionuclide angiography, but this approach has been limited by fast radiotracer diffusion. Ultra-sensitive digital PET systems can produce high-quality images within 3-s acquisition times. This study determined whether digital PET/CT accurately measured LVEF in an anthropomorphic heart phantom under conditions mimicking radiotracer first-pass into the cardiac cavities.MethodsHeart phantoms in end-diastole and end-systole were 3D-printed from a patient’s MRI dataset. Reference left ventricle end-diastole volume (EDV), end-systole volume (ESV), and LVEF were determined by phantom weights before/after water filling. PET/CT (3-s acquisitions), MRI, and planar and SPECT ERNA were performed. EDV, ESV, and/or LVEF were measured by manual and automated cardiac cavity delineation, using clinical segmentation softwares. LVEF was also measured from PET images converted to 2D “pseudo-planar” images along the short axis and horizontal long axis. LVEF was also calculated for planar ERNA images. All LVEF, ESV and EDV values were compared to the reference values assessed by weighing.ResultsManually calculated 3D-PET-CT-based EDV, ESV, and LVEF were close to MRI and reference values. Automated calculations on the 3D-PET-CT dataset were unreliable, suggesting that the SPECT-based tool used for this calculation is not well adapted for PET acquisitions. Manual and automated LVEF estimations from “pseudo-planar” PET images were very close/identical to MRI and reference values.ConclusionsFirst-pass “pseudo-planar” PET may be a promising method for estimating LVEF, easy to use in clinical practice. Processing 3D PET images is also a valid method but to date suffers from a lack of well-suited software for automated LV segmentation.

Ground-truth-free deep learning for artefacts reduction in 2D radial cardiac cine MRI using a synthetically generated dataset

In this work, we consider the task of image reconstruction in 2D radial cardiac cine MRI using deep learning (DL)-based regularization. As the regularization is achieved by employing an image-prior predicted by a pre-trained convolutional neural network (CNN), the quality of the image-prior is of essential importance. The achievable performance of any DL-based method is limited by the amount and the quality of the available training data. For fast dynamic processes, obtaining good-quality MR data is challenging because of technical and physiological reasons. In this work, we try to overcome these problems by a transfer-learning approach which is motivated by a previously presented DL-method (XT,YT U-Net). There, instead of training the network on the whole 2D dynamic images, it is trained on 2D spatio-temporal profiles (xt,yt-slices) which show the temporal changes of the imaged object. Therefore, for the training and test data, it is more important that their spatio-temporal profiles share similar local features rather than being images of the same anatomy. This allows us to equip arbitrary data with simulated motion that resembles the cardiac motion and use it as training data. By doing so, it is possible to train a CNN which is applicable to cardiac cine MR data without using ground-truth cine MR images for training. We demonstrate that combining XT,YT U-Net with the proposed transfer-learning strategy delivers comparable performance to CNNs trained on cardiac cine MR images and in some cases even qualitatively surpasses these. Additionally, the transfer-learning strategy was investigated for a 2D and 3D U-Net. The images processed by the the CNNs were used as image-priors in the CNN-regularized iterative reconstruction. The XT,YT U-Net yielded visibly better results than the 2D U-Net and slightly better results than the 3D U-Net when used in combination with the presented transfer learning-strategy.

Cine Cardiac MRI Motion Artifact Reduction Using a Recurrent Neural Network.

Cine cardiac magnetic resonance imaging (MRI) is widely used for the diagnosis of cardiac diseases thanks to its ability to present cardiovascular features in excellent contrast. As compared to computed tomography (CT), MRI, however, requires a long scan time, which inevitably induces motion artifacts and causes patients' discomfort. Thus, there has been a strong clinical motivation to develop techniques to reduce both the scan time and motion artifacts. Given its successful applications in other medical imaging tasks such as MRI super-resolution and CT metal artifact reduction, deep learning is a promising approach for cardiac MRI motion artifact reduction. In this paper, we propose a novel recurrent generative adversarial network model for cardiac MRI motion artifact reduction. This model utilizes bi-directional convolutional long short-term memory (ConvLSTM) and multi-scale convolutions to improve the performance of the proposed network, in which bi-directional ConvLSTMs handle long-range temporal features while multi-scale convolutions gather both local and global features. We demonstrate a decent generalizability of the proposed method thanks to the novel architecture of our deep network that captures the essential relationship of cardiovascular dynamics. Indeed, our extensive experiments show that our method achieves better image quality for cine cardiac MRI images than existing state-of-the-art methods. In addition, our method can generate reliable missing intermediate frames based on their adjacent frames, improving the temporal resolution of cine cardiac MRI sequences.

Super-Resolution of Cardiac MR Cine Imaging using Conditional GANs and Unsupervised Transfer Learning.

High-resolution (HR), isotropic cardiac Magnetic Resonance (MR) cine imaging is challenging since it requires long acquisition and patient breath-hold times. Instead, 2D balanced steady-state free precession (SSFP) sequence is widely used in clinical routine. However, it produces highly-anisotropic image stacks, with large through-plane spacing that can hinder subsequent image analysis. To resolve this, we propose a novel, robust adversarial learning super-resolution (SR) algorithm based on conditional generative adversarial nets (GANs), that incorporates a state-of-the-art optical flow component to generate an auxiliary image to guide image synthesis. The approach is designed for real-world clinical scenarios and requires neither multiple low-resolution (LR) scans with multiple views, nor the corresponding HR scans, and is trained in an end-to-end unsupervised transfer learning fashion. The designed framework effectively incorporates visual properties and relevant structures of input images and can synthesise 3D isotropic, anatomically plausible cardiac MR images, consistent with the acquired slices. Experimental results show that the proposed SR method outperforms several state-of-the-art methods both qualitatively and quantitatively. We show that subsequent image analyses including ventricle segmentation, cardiac quantification, and non-rigid registration can benefit from the super-resolved, isotropic cardiac MR images, to produce more accurate quantitative results, without increasing the acquisition time. The average Dice similarity coefficient (DSC) for the left ventricular (LV) cavity and myocardium are 0.95 and 0.81, respectively, between real and synthesised slice segmentation. For non-rigid registration and motion tracking through the cardiac cycle, the proposed method improves the average DSC from 0.75 to 0.86, compared to the original resolution images.

Deep learning based fully automatic segmentation of the left ventricular endocardium and epicardium from cardiac cine MRI.

The segmentation of cardiac medical images is a crucial step for calculating clinical indices such as wall thickness, ventricular volume, and ejection fraction. In this study, we introduce a method named LsUnet that combines multi-channel, fully convolutional neural network, and annular shape level-set methods for efficiently segmenting cardiac cine magnetic resonance (MR) images. In this method, the multi-channel deep learning algorithm is applied to train the segmentation task to extract the left ventricle (LV) endocardial and epicardial contours. Next, the segmentation contours from the multi-channel deep learning method are incorporated into a level-set formulation, which is dedicated explicitly to detecting annular shapes to assure the segmentation's accuracy and robustness. The proposed automatic approach was evaluated on 95 volumes (total 1,076 slices, ~80% as for training datasets, ~20% 2D as for testing datasets). This combined multi-channel deep learning and annular shape level-set segmentation method achieved high accuracy with average Dice values reaching 92.15% and 95.42% for LV endocardium and epicardium delineation, respectively, in comparison to the reference standard (the manual segmentation). A novel method for fully automatic segmentation of the LV endocardium and epicardium from different MRI datasets is presented. The proposed workflow is accurate and robust compared to the reference and other state-of-the-art methods.

Assessment of myocardial deformation with CMR: a comparison with ultrasound speckle tracking.

Myocardial deformation integrated with cardiac dimensions provides a comprehensive assessment of cardiac function, which has proven useful to differentiate cardiac pathology from physiological adaptation to situations such as chronic intensive training. Feature tracking (FT) can measure myocardial deformation from cardiac magnetic resonance (CMR) cine sequences; however, its accuracy is not yet fully validated. Our aim was to compare the accuracy and reproducibility of FT with speckle tracking echocardiography (STE) in highly trained endurance athletes. Ninety-three endurance athletes (> 12-h training/week during the last 5 years, 52% male, 35 ± 5.1 years old) and 72 age-matched controls underwent resting CMR and transthoracic echocardiography to assess biventricular exercise-induced remodeling and biventricular global longitudinal strain (GLS) by CMR-FT and STE. Strain values were significantly lower when assessed by CMR-FT compared to STE (p < 0.001), with good reproducibility for the left ventricle (bias = 3.94%, limit of agreement [LOA] = ± 4.27 %) but wider variability for right ventricle strain. Strain values by both techniques proportionally decreased with increasing ventricular volumes, potentially depicting the functional biventricular reserve that characterizes athletes' hearts. Biventricular longitudinal strain values were lower when assessed by FT as compared to STE. Both methods were statistically comparable when measuring LV strain but not RV strain. These differences might be justified by the lower in-plane spatial and temporal resolution of FT, which is particularly relevant for the complex anatomy of the RV. • Strain values were significantly lower when assessed by FT as compared to STE, which was expected due to the lower in-plane spatial and temporal resolution of FT versus STE. • Both methods were statistically comparable when measuring LV strain but not for RV strain analysis. • Characterizing the normal ranges and reproducibility of strain metrics by FT is an important step toward its clinical applicability, since it can be assessed offline and applied to routinely acquired cine CMR images.

Automated left and right ventricular chamber segmentation in cardiac magnetic resonance images using dense fully convolutional neural network

Background and objectiveSegmentation of the left ventricular (LV) myocardium (Myo) and RV endocardium on cine cardiac magnetic resonance (CMR) images represents an essential step for cardiac-function evaluation and diagnosis. In order to have a common reference for comparing segmentation algorithms, several CMR image datasets were made available, but in general they do not include the most apical and basal slices, and/or gold standard tracing is limited to only one of the two ventricles, thus not fully corresponding to real clinical practice. Our aim was to develop a deep learning (DL) approach for automated segmentation of both RV and LV chambers from short-axis (SAX) CMR images, reporting separately the performance for basal slices, together with the applied criterion of choice. MethodA retrospectively selected database (DB1) of 210 cine sequences (3 pathology groups) was considered: images (GE, 1.5 T) were acquired at Centro Cardiologico Monzino (Milan, Italy), and end-diastolic (ED) and end-systolic frames (ES) were manually segmented (gold standard, GS). Automatic ED and ES RV and LV segmentation were performed with a U-Net inspired architecture, where skip connections were redesigned introducing dense blocks to alleviate the semantic gap between the U-Net encoder and decoder. The proposed architecture was trained including: A) the basal slices where the Myo surrounded the LV for at least the 50% and all the other slice; B) all the slices where the Myo completely surrounded the LV. To evaluate the clinical relevance of the proposed architecture in a practical use case scenario, a graphical user interface was developed to allow clinicians to revise, and correct when needed, the automatic segmentation. Additionally, to assess generalizability, analysis of CMR images obtained in 12 healthy volunteers (DB2) with different equipment (Siemens, 3T) and settings was performed. ResultsThe proposed architecture outperformed the original U-Net. Comparing the performance on DB1 between the two criteria, no significant differences were measured when considering all slices together, but were present when only basal slices were examined. Automatic and manually-adjusted segmentation performed similarly compared to the GS (bias±95%LoA): LVEDV -1±12 ml, LVESV -1±14 ml, RVEDV 6±12 ml, RVESV 6±14 ml, ED LV mass 6±26 g, ES LV mass 5±26 g). Also, generalizability showed very similar performance, with Dice scores of 0.944 (LV), 0.908 (RV) and 0.852 (Myo) on DB1, and 0.940 (LV), 0.880 (RV), and 0.856 (Myo) on DB2. ConclusionsOur results support the potential of DL methods for accurate LV and RV contours segmentation and the advantages of dense skip connections in alleviating the semantic gap generated when high level features are concatenated with lower level feature. The evaluation on our dataset, considering separately the performance on basal and apical slices, reveals the potential of DL approaches for fast, accurate and reliable automated cardiac segmentation in a real clinical setting.

Joint Deep Learning Framework for Image Registration and Segmentation of Late Gadolinium Enhanced MRI and Cine Cardiac MRI.

Late gadolinium enhanced (LGE) cardiac magnetic resonance (CMR) imaging, the current benchmark for assessment of myocardium viability, enables the identification and quantification of the compromised myocardial tissue regions, as they appear hyper-enhanced compared to the surrounding, healthy myocardium. However, in LGE CMR images, the reduced contrast between the left ventricle (LV) myocardium and LV blood-pool hampers the accurate delineation of the LV myocardium. On the other hand, the balanced-Steady State Free Precession (bSSFP) cine CMR imaging provides high resolution images ideal for accurate segmentation of the cardiac chambers. In the interest of generating patient-specific hybrid 3D and 4D anatomical models of the heart, to identify and quantify the compromised myocardial tissue regions for revascularization therapy planning, in our previous work, we presented a spatial transformer network (STN) based convolutional neural network (CNN) architecture for registration of LGE and bSSFP cine CMR image datasets made available through the 2019 Multi-Sequence Cardiac Magnetic Resonance Segmentation Challenge (MS-CMRSeg). We performed a supervised registration by leveraging the region of interest (RoI) information using the manual annotations of the LV blood-pool, LV myocardium and right ventricle (RV) blood-pool provided for both the LGE and the bSSFP cine CMR images. In order to reduce the reliance on the number of manual annotations for training such network, we propose a joint deep learning framework consisting of three branches: a STN based RoI guided CNN for registration of LGE and bSSFP cine CMR images, an U-Net model for segmentation of bSSFP cine CMR images, and an U-Net model for segmentation of LGE CMR images. This results in learning of a joint multi-scale feature encoder by optimizing all three branches of the network architecture simultaneously. Our experiments show that the registration results obtained by training 25 of the available 45 image datasets in a joint deep learning framework is comparable to the registration results obtained by stand-alone STN based CNN model by training 35 of the available 45 image datasets and also shows significant improvement in registration performance when compared to the results achieved by the stand-alone STN based CNN model by training 25 of the available 45 image datasets.

Aortic root longitudinal strain by speckle-tracking echocardiography predicts progressive aortic root dilation in Marfan syndrome patients

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Spanish Ministry of Science, Innovation and Universities; Instituto de Salud Carlos III Introduction In Marfan syndrome (MFS) patients reduced longitudinal strain of the ascending aorta (AAo) as measured by applying feature-tracking on cine cardiac magnetic resonance (CMR) images predicts aortic root dilation and aortic events during the follow-up. Speckle-tracking is well established for cardiac deformation assessment but proximal aorta applications are challenging due to limited wall thickness and substantial cardiac motion. Moreover, echocardiography is widely used in the clinical assessment aortic diseases. Purpose We aimed to test a speckle-tracking tool for root longitudinal strain analysis in terms of comparison with CMR-derived AAo longitudinal strain and reproducibility and as predictor of dilation in MFS patients. Methods Thirty-five MFS patients diagnosed by original GHENT criteria, with maximum aortic root diameter of 45 mm and free from previous aortic dissection or cardiac/aortic surgery and non-severe aortic regurgitation were consecutive enrolled and followed-up. CMR and echocardiography were performed less than 2 months apart. Baseline and final aortic root diameter were measured on CMR images. To quantify aortic root cyclic elongation by echocardiography, two regions of interests were manually created covering both walls in a parasternal long-axis view and tracked along the cardiac cycle (Figure 1). Longitudinal strain was computed as the average of maximum increase in relative distance of several sub-regions covering both walls. CMR-derived AAo longitudinal strain was available in 29 patients. Intra-observer reproducibility was tested in 15 patients via intraclass correlation coefficient (ICC) for single-rater absolute agreement. Results Aortic root longitudinal strain by echocardiography was mildly related to CMR-derived AAo longitudinal strain (R = 0.27) and was larger compared to CMR-derived values (16.2 ± 6.0 vs 11.3 ± 4.3). Reproducibility was high, with ICC of 0.811, R = 0.802, p &amp;lt; 0.001. After a mean follow up of 76 ± 13 months, aortic root diameter grew in 20 patients with a rate of 0.29± 0.24 mm/year. Overall mean growth-rate was 0.87 ± 0.33 mm/year. In multivariable analysis corrected for age and baseline aortic root diameter, baseline longitudinal strain by echocardiography was independently and inversely related to progressive dilation (p = 0.033). Conclusions The measurement of aortic root longitudinal strain by speckle-tracking echocardiography is feasible. Aortic root longitudinal strain is an independent predictor of progressive dilation in MFS patients. This may permit the improvement of risk-stratification in aortic diseases in large scale studies. Abstract Figure 1

The interplay of systemic atrial and ventricular function in patients with transposition of the great arteries after arterial switch operation

Abstract Funding Acknowledgements Type of funding sources: None. Background/Introduction: The role of systemic atrial (LA) function and its interplay with the systemic ventricle (LV) in patients with transposition of the great arteries (TGA) after arterial switch operation (ASO) is currently not well defined. Purpose Investigate the role of LA function and its interplay with the LV. Methods Between 2007 and 2018 patients with TGA corrected by ASO and control patients of comparable age, sex, height but without known cardiovascular disease were included in this prospective single centre analysis of 3 Tesla cardiac magnetic resonance cine imaging studies of atrial and ventricular function. Additionally, in TGA patients extracellular volume fraction (ECV) was assessed by T1 mapping and global longitudinal strain (GLS) by feature tracking. Results Overall, 81 TGA (16.3 [IQR 10.5-21.2] years, 32% female) and 30 control patients (11.9 [IQR 8.7-22.5] years, 63% female) were included in the current analysis. TGA patients had smaller LA maximum volume index, and reduced total atrial and conduit volume emptying. This resulted in a reduced LA reservoir and conduit function compared to controls. In TGA, a higher LA active/conduit and active/total ratio indicated impaired passive filling of the LV. (Table 1) The median ECV was 28.3% (IQR 25.8-33.9) and the median GLS -24.2% (IQR -28.6- -20.4) in TGA patients. LA reservoir (Fig. 1A), conduit and contractile function showed a negative correlation with GLS (r=-0.470, p &amp;lt; 0.001, r=-0.270, p = 0.022 and r=-0.257, p = 0.030, respectively). Interestingly, the LA active/conduit ratio (Fig. 1B) and the active/total ratio showed a positive correlation with ECV (r = 0.418, p = 0.002 and r = 0.339, p = 0.013, respectively). Conclusion Impaired LA function is frequent among patients with TGA following ASO. The impairment of LA function is linked to both LV dysfunction and fibrosis. Table 1 Variable TGA (n = 81) Control (n = 30) p-value LAVI max.[ml/m&amp;sup2;] 37.8 (29.9-42.9) 42.02 (38.3-51.6) 0.002 LAVI total emptying [ml/m&amp;sup2;] 16.8 (13.7-20.8) 24.3 (20.6-29.4) &amp;lt;0.001 LAVI conduit emptying [ml/m&amp;sup2;] 10.3 (8.1-13.9) 17.5 (15.2-20.8) &amp;lt;0.001 LAVI active [ml/m&amp;sup2;] 6.7 (4.1-8.6) 6.4 (5.0-7.9) 0.984 LA-EF-reservoir [%] 46.6 (42.0-50.5) 57.2 (51.2-60.1) &amp;lt;0.001 LA-EF-conduit [%] 28.1 (23.4-34.3) 42.5 (35.4-45.8) &amp;lt;0.001 LA-EF-active [%] 24.11 (17.9-29.1) 25.7 (21.5-29.7) 0.339 LA active/conduit ratio 0.60 (0.41-0.90) 0.37 (0.28-0.44) &amp;lt;0.001 LA active/total ratio 0.38 (0.29-0.47) 0.27 (0.22-0.30) &amp;lt;0.001 Differences in left atrial function between TGA and control patients Abstract Figure. Fig. 1A and B

Feasibility of single-shot compressed sensing cine imaging for analysis of left ventricular function and strain in cardiac MRI

To compare single-shot compressed sensing (CS) cine imaging with conventional segmented cine imaging for reliable quantification of left ventricular (LV) volume and strain assessment during cardiac magnetic resonance imaging (CMRI). Thirty-seven participants underwent both single-shot CS and conventional segmented cines that covered the entire LV. LV volumetric and strain values were obtained. LV volumes, global strain, the standard deviation of time to peak strain (SD-TPS) in the radial, longitudinal, and circumferential directions were compared using the Student's t-test and intraclass correlation coefficient (ICC). Interobserver and intra-observer variabilities of the LV strain values for the two cines method were determined using ICC. Single-shot CS cine-derived LV volumes and myocardial mass measurements correlated strongly with segmented cines (ICC >0.798) and minor systematic end-systolic volume overestimations resulting in ejection fraction underestimations. Single-shot CS cine-derived global strain and SD-TPS were poorly to moderately correlated with segmented cines (ICC from 0.045-0.706). All global strain values derived from single-shot CS cines were underestimated compared with segmented cine-derived values; however, no significant differences in radial and longitudinal SD-TPS between the two cines were found. Among the patient-related factors, heart rate was a strong predictive factor of global longitudinal strain underestimations (p=0.039) in the CS cines. Inter- and intra-observer LV strain variabilities derived from CS and segmented cines were good to excellent. Single-shot CS cine CMRI is feasible for the quantitative assessment of LV function. Currently, strain values derived from the two techniques are not interchangeable.

Reliability and reproducibility of cardiac MRI quantification of peak exercise function with long-axis views.

The conventional approach to cardiac magnetic resonance (CMR) involving breath holds, electrocardiography-gating, and acquisition of a short-axis (SAX) image stack, introduces technical and logistical challenges for assessing exercise left ventricular (LV) function. Real-time, free-breathing CMR acquisition of long-axis (LAX) images overcomes these issues and also enables assessment of global longitudinal strain (GLS). We evaluated the reliability of a free-breathing LAX approach compared to the standard SAX approach and the reproducibility of free-breathing LAX. LV SAX (contiguous stack) and LAX (two-chamber and four-chamber) 3T CMR cine images were acquired four times within one scan in 32 women with cardiovascular risk factors (56±10 years, 28±4 kg/m2) as follows: 1) resting, gated-segmented, end-expiration breath-hold; 2) resting, real-time, free-breathing; 3) test-retest set of resting, real-time, free-breathing; 4) peak exercise (incremental-to-maximum, in-magnet, stepper test), real-time, free-breathing. A second scan was performed within one week in a subset (n = 5) to determine reproducibility of peak exercise measures. Reliability and agreement of the free-breathing LAX approach with the conventional SAX approach were assessed by intraclass correlation coefficient (ICC) and Bland-Altman plots, respectively. Normal control GLS reserve was also acquired in a separate set of 12 young, healthy control women (25±4 years, 22±2 kg/m2) for comparison. Comparisons of LV volumes and function among all techniques at rest had good-to-excellent reliability (ICC = 0.80–0.96), and excellent reliability between peak exercise free-breathing LAX and SAX evaluations (ICC = 0.92–0.96). Higher resting heart rates with free-breathing acquisitions compared to breath-hold (mean difference, limits of agreement: 5, 1–12 beats per minute) reduced reliability for cardiac output (ICC = 0.67–0.79). Reproducibility of the free-breathing LAX approach was good-to-excellent at rest and peak exercise (ICC = 0.74–0.99). GLS exercise reserve was impaired in older women at cardiovascular risk compared to young healthy women (-4.7±2.3% vs -7.4±2.1%, p = 0.001). Real-time, free-breathing CMR with LAX evaluation provides a reliable and reproducible method to assess rest and peak exercise cardiac function, including GLS.

Network slimming for compressed‐sensing cardiac cine MRI

The Unet is attempted to be slimmed for compressed-sensing cardiac cine magnetic resonance imaging. Despite the excellent performance of the U-net, its heavy structure and long training time restrict its applications in an environment with limited computational resources. We slimmed the U-net by changing the multiple convolutions to single convolution and the transpose convolution to upsampling and by adopting fewer layers, without performance degradation. The number of trainable weights and training time of the slimmed network was reduced by 87.9% and 48.1%, respectively. The proposed network showed improved performance with a 1.45% reduction in the normalised mean square error.

Automatic cardiac cine MRI segmentation and heart disease classification

Cardiac cine magnetic resonance imaging (MRI) continues to be recognized as an established modality for non-invasive assessment of the function and structure of the cardiovascular system. Making full use of fully convolutional neural networks CNNs ability to operate pixel-wise classification, cine MRI sequences can be segmented and volumetric features of three key heart structures are computed for disease prediction. The three key heart structures are the left ventricle cavity, right ventricle cavity and the left ventricle myocardium. In this paper, we suggest an automated pipeline for both cardiac segmentation and diagnosis. The study was conducted on a dataset of 150 patients from Dijon Hospital in the context of the post-2017 Medical Image Computing and Computer Assisted Intervention MICCAI, Automated Cardiac Diagnosis Challenge (ACDC). The challenge consists in two phases: (i) a segmentation contest, where performance is evaluated on dice overlap coefficient and Hausdorff distance metrics, and a (ii) diagnosis contest for heart disease classification. For this aim, we propose the use of a deep learning based network for segmentation of the three key cardiac structures within short-axis cine MRI sequences and a classifier ensemble for heart disease classification. The deep learning segmentation network is a UNet fully convolutional neural network variant with fewer trainable parameters. The classifier ensemble consists in combining three classifiers, namely a multilayer perceptron, a random forest and a support vector machine. Before feeding the segmentation network, a preliminary step consists in localizing heart region and cropping input images to a restricted region of interest (ROI). This is achieved by a signal processing based approach and aims at reducing multi-class imbalance and computational load. We achieved nearly state of the art accuracy performances for both the segmentation and disease classification challenges. Reporting a mean dice overlap coefficient of 0.92 for the three cardiac structures segmentation, along with good limits of agreement for the various derived clinical indices, leading to an accuracy of 0.92 for the disease classification on unseen data.

CNN-Based Cardiac Motion Extraction to Generate Deformable Geometric Left Ventricle Myocardial Models from Cine MRI.

Patient-specific left ventricle (LV) myocardial models have the potential to be used in a variety of clinical scenarios for improved diagnosis and treatment plans. Cine cardiac magnetic resonance (MR) imaging provides high resolution images to reconstruct patient-specific geometric models of the LV myocardium. With the advent of deep learning, accurate segmentation of cardiac chambers from cine cardiac MR images and unsupervised learning for image registration for cardiac motion estimation on a large number of image datasets is attainable. Here, we propose a deep leaning-based framework for the development of patient-specific geometric models of LV myocardium from cine cardiac MR images, using the Automated Cardiac Diagnosis Challenge (ACDC) dataset. We use the deformation field estimated from the VoxelMorph-based convolutional neural network (CNN) to propagate the isosurface mesh and volume mesh of the end-diastole (ED) frame to the subsequent frames of the cardiac cycle. We assess the CNN-based propagated models against segmented models at each cardiac phase, as well as models propagated using another traditional nonrigid image registration technique. Additionally, we generate dynamic LV myocardial volume meshes at all phases of the cardiac cycle using the log barrier-based mesh warping (LBWARP) method and compare them with the CNN-propagated volume meshes.

An image registration framework to estimate 3D myocardial strains from cine cardiac MRI in mice.

Accurate and efficient quantification of cardiac motion offers promising biomarkers for non-invasive diagnosis and prognosis of structural heart diseases. Cine cardiac magnetic resonance imaging remains one of the most advanced imaging tools to provide image acquisitions needed to assess and quantify in-vivo heart kinematics. The majority of cardiac motion studies are focused on human data, and there remains a need to develop and implement an image-registration pipeline to quantify full three-dimensional (3D) cardiac motion in mice where ideal image acquisition is challenged by the subject size and heart rate and the possibility of traditional tagged imaging is hampered. In this study, we used diffeomorphic image registration to estimate strains in the left ventricular wall in two wild-type mice and one diabetic mouse. Our pipeline resulted in a continuous and fully 3D strain map over one cardiac cycle. The estimation of 3D regional and transmural variations of strains is a critical step towards identifying mechanistic biomarkers for improved diagnosis and phenotyping of structural left heart diseases including heart failure with reduced or preserved ejection fraction.

A Convolutional Neural Network-based Deformable Image Registration Method for Cardiac Motion Estimation from Cine Cardiac MR Images.

In this work, we describe an unsupervised deep learning framework featuring a Laplacian-based operator as smoothing loss for deformable registration of 3D cine cardiac magnetic resonance (CMR) images. Before registration, the input 3D images are corrected for slice misalignment by segmenting the left ventricle (LV) blood-pool, LV myocardium and right ventricle (RV) blood-pool using a U-Net model and aligning the 2D slices along the center of the LV blood-pool. We conducted experiments using the Automated Cardiac Diagnosis Challenge (ACDC) dataset. We used the registration deformation field to warp the manually segmented LV blood-pool, LV myocardium and RV blood-pool labels from end-diastole (ED) frame to the other frames in the cardiac cycle. We achieved a mean Dice score of 94.84%, 85.22% and 84.36%, and Hausdorff distance (HD) of 2.74 mm, 5.88 mm and 9.04 mm, for the LV blood-pool, LV myocardium and RV blood-pool, respectively. We also introduce a pipeline to estimate patient tractography using the proposed CNN-based cardiac motion estimation.

Abstract 14478: Left Ventricular (LV) Determinants of Left Atrial (LA) Remodeling in Ischemic Cardiomyopathy (iCM): From the Allogeneic Heart Stem Cells to Achieve Myocardial Regeneration (ALLSTAR) Double Blind Placebo Controlled Trial

Introduction: ALLSTAR, a trial of cardiosphere-derived cells in iCM, phenotyped the subjects by cardiac magnetic resonance (CMR) at baseline, and 6- and 12- months post-randomization. The overall CMR dataset provides an opportunity to study temporal remodeling of the LA in the context of iCM-related LV alterations. Hypothesis: We hypothesize that LA volumetric and functional remodeling in iCM are coupled to LV determinants. Methods: LA volumes (LAV; maximum and minimum) were obtained from 4- and 2-chamber cine CMR images. LA functional parameters included total, passive and active LA emptying fractions and global longitudinal strain (GLS). LV parameters included end-diastolic and end-systolic volumes (indexed to body surface area; LVEDVi and LVESVi), as well as LV ejection fraction (EF), infarct size, GLS and end-diastolic mass. Six- and 12-month differences from the baseline CMR study were calculated for LA and LV parameters. Linear regression models were used to evaluate univariate associations between LA and LV parameters (at baseline, 6-months, 12-months and their differences [Δ]), independent of treatment effect (which is the subject of ongoing analysis). Results: 101 participants had 6-month follow up CMR studies; 57 had 12-month studies. Baseline means(SD): age 53.9(10.9) years; 87(86.1%) male; LVEF 39.7%(7.0%); infarct size 22.7%(5.4%). In univariate analyses at 6 months, ΔLVEDVi and ΔLVESVi were both significantly associated with ΔLAVimax and ΔLAVimin, but not with LA functional parameters (Table). At 12 months, ΔLVEDVi remained significantly associated with ΔLAVimax, and ΔLVESVi with ΔLAVimin. No association with LA functional parameters was seen. LVEF, infarct size, GLS and end-diastolic mass were not associated with any volumetric or functional LA parameters (not shown). Conclusions: In conclusion, in this cohort of iCM, pathologic dilatation of the LV appears coupled to volumetric LA remodeling, but not other functional LA parameters.