An iterative multi‐path fully convolutional neural network for automatic cardiac segmentation in cine MR images

Abstract

Purpose Segmentation of the left ventricle (LV), right ventricle (RV) cavities and the myocardium (MYO) from cine cardiac magnetic resonance (MR) images is an important step for diagnosis and monitoring cardiac diseases. Spatial context information may be highly beneficial for segmentation performance improvement. To this end, this paper proposes an iterative multi‐path fully convolutional network (IMFCN) to effectively leverage spatial context for automatic cardiac segmentation in cine MR images. Methods To effectively leverage spatial context information, the proposed IMFCN explicitly models the interslice spatial correlations using a multi‐path late fusion strategy. First, the contextual inputs including both the adjacent slices and the already predicted mask of the above adjacent slice are processed by independent feature‐extraction paths. Then, an atrous spatial pyramid pooling (ASPP) module is employed at the feature fusion process to combine the extracted high‐level contextual features in a more effective way. Finally, deep supervision (DS) and batch‐wise class re‐weighting mechanism are utilized to enhance the training of the proposed network. Results The proposed IMFCN was evaluated and analyzed on the MICCAI 2017 automatic cardiac diagnosis challenge (ACDC) dataset. On the held‐out training dataset reserved for testing, our method effectively improved its counterparts that without spatial context and that with spatial context but using an early fusion strategy. On the 50 subjects test dataset, our method achieved Dice similarity coefficient of 0.935, 0.920, and 0.905, and Hausdorff distance of 7.66, 12.10, and 8.80 mm for LV, RV, and MYO, respectively, which are comparable or even better than the state‐of‐the‐art methods of ACDC Challenge. In addition, to explore the applicability to other datasets, the proposed IMFCN was retrained on the Sunnybrook dataset for LV segmentation and also produced comparable performance to the state‐of‐the‐art methods. Conclusions We have presented an automatic end‐to‐end fully convolutional architecture for accurate cardiac segmentation. The proposed method provides an effective way to leverage spatial context in a two‐dimensional manner and results in precise and consistent segmentation results.

Fully automated quantification of left ventricular volumes and function in cardiac MRI: clinical evaluation of a deep learning-based algorithm

“…However, even with these corrections the expert corrected automated analysis took less than 2 min on average-significantly less than the manual analysis, which on average took 3.5 or 9 min, depending on level of experience. Several previous publications focus on the technical aspects of developing and validating deep learning-based algorithms for fully automated volumetric analysis of cardiac MRI [12,[16][17][18][19]. Bai and colleagues describe the development and validation of a fully automated method capable of segmenting the volume of both ventricles and atria based on a multi-layered fully convolutional network [12].…”

Fully Automatic Left Ventricle Segmentation Using Bilateral Lightweight Deep Neural Network

“…For LV segmentation in CT images, an FCN-based architecture is utilized with pre-trained weights of VGG [ 21 ]. Another revised version of FCN with different loss functions is analyzed in [ 22 ] and an iterative multi-path FCN (IMFCN) segmentation model is proposed, which segments the LV and RV from MRI images. However, the most significant weakness of FCN is the loss of spatial information.…”

An Adaptive Generative Adversarial Network for Cardiac Segmentation from X-ray Chest Radiographs

“…It is also the first learning algorithm to be successfully used to train a multilayer network. A convolutional neural network includes an input layer, convolutional layers, downsampling layers (also known as pooling layers), connected layers and an output layer, as shown in Figure 1 [26].…”

“…It is also the first learning algorithm to be successfully used to train a multilayer network. A convolutional neural network includes an input layer, convolutional layers, downsampling layers (also known as pooling layers), connected layers and an output layer, as shown in Figure 1 [26]. After image convolution is completed as shown above, to reduce the dimensionality of the feature map and maintain its feature size invariant to a certain extent, the feature map needs to be downsampled in accordance with certain rules.…”

“…There is no restriction on the specific structures of the generator and discriminator. Due to the remarkable achievements of deep neural networks in image processing [26], the feature extractor and discriminator in this paper are both designed on the basis of neural network models. At the same time, to better retain the image details, the features extracted in two dimensions are also combined to make the segmentation more accurate after adaptive selection.…”