Search

Quantitative myocardial contrast echocardiography (MCE) has been proved to be valuable in detecting myocardial ischemia. During quantitative MCE analysis, myocardial segmentation is a critical step in determining accurate region of interests (ROIs). However, traditional myocardial segmentation mainly relies on manual tracing of myocardial contours, which is time-consuming and laborious. To solve this problem, we propose a fully automatic myocardial segmentation framework that can segment myocardial regions in MCE accurately without human intervention. A total of 100 patients’ MCE sequences were divided into a training set and a test set according to a 7: 3 proportion for analysis. We proposed a bi-directional training schema, which incorporated temporal information of forward and backward direction among frames in MCE sequences to ensure temporal consistency by combining convolutional neural network with recurrent neural network. Experiment results demonstrated that compared with a traditional segmentation model (U-net) and the model considering only forward temporal information (U-net + forward), our framework achieved the highest segmentation precision in Dice coefficient (U-net vs U-net + forward vs our framework: 0.78 ± 0.07 vs 0.79 ± 0.07 vs 0.81 ± 0.07, p

Most existing deep learning-based frameworks for image segmentation assume that a unique ground truth is known and can be used for performance evaluation. This is true for many applications, but not all. Myocardial segmentation of Myocardial Contrast Echocardiography (MCE), a critical task in automatic myocardial perfusion analysis, is an example. Due to the low resolution and serious artifacts in MCE data, annotations from different cardiologists can vary significantly, and it is hard to tell which one is the best. In this case, how can we find a good way to evaluate segmentation performance and how do we train the neural network? In this paper, we address the first problem by proposing a new extended Dice to effectively evaluate the segmentation performance when multiple accepted ground truth is available. Then based on our proposed metric, we solve the second problem by further incorporating the new metric into a loss function that enables neural networks to flexibly learn general features of myocardium. Experiment results on our clinical MCE data set demonstrate that the neural network trained with the proposed loss function outperforms those existing ones that try to obtain a unique ground truth from multiple annotations, both quantitatively and qualitatively. Finally, our grading study shows that using extended Dice as an evaluation metric can better identify segmentation results that need manual correction compared with using Dice.