Cardiac magnetic resonance imaging (CMRI) provides high resolution images ideal for assessing cardiac function and diagnosis of cardiovascular diseases. To assess cardiac function, estimation of ejection fraction, ventricular volume, mass and stroke volume are crucial, and the segmentation of left ventricle from CMRI is the first critical step. Fully convolutional neural network architectures have proved to be very efficient for medical image segmentation, with U-Net inspired architecture as the current state-of-the-art. Generative adversarial networks (GAN) inspired architectures have recently gained popularity in medical image segmentation with one of them being SegAN, a novel end-to-end adversarial neural network architecture. In this paper, we investigate SegAN with three different types of U-Net inspired architectures for left ventricle segmentation from cardiac MRI data. We performed our experiments on the 2017 ACDC segmentation challenge dataset. Our results show that the performance of U-Net architectures is better when trained in the SegAN framework than when trained stand-alone. The mean Dice scores achieved for three different U-Net architectures trained in the SegAN framework was on the order of 93.62%, 92.49% and 94.57%, showing a significant improvement over their Dice scores following stand-alone training-92.58%), 91.46% and 93.81%, respectively.

No abstract

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.

No abstract

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 patientspecific 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.