Automatic Myocardial Segmentation by Using A Deep Learning Network in Cardiac MRI
This work addresses the need for efficient and accurate myocardial segmentation in cardiac MRI, which is crucial for diagnosing conditions like mitral regurgitation and myocarditis, but it is incremental as it builds on existing deep learning techniques with specific enhancements.
The authors tackled the problem of automating myocardial segmentation in cardiac MRI by proposing a deep learning network with improvements like the Jaccard distance as an objective function, residual learning, and batch normalization, resulting in a method that outperforms previous approaches and processes a volume in under 22 seconds.
Cardiac function is of paramount importance for both prognosis and treatment of different pathologies such as mitral regurgitation, ischemia, dyssynchrony and myocarditis. Cardiac behavior is determined by structural and functional features. In both cases, the analysis of medical imaging studies requires to detect and segment the myocardium. Nowadays, magnetic resonance imaging (MRI) is one of the most relevant and accurate non-invasive diagnostic tools for cardiac structure and function. In this work we propose to use a deep learning technique to assist the automatization of myocardial segmentation in cardiac MRI. We present several improvements to previous works in this paper: we propose to use the Jaccard distance as optimization objective function, we integrate a residual learning strategy into the code, and we introduce a batch normalization layer to train the fully convolutional neural network. Our results demonstrate that this architecture outperforms previous approaches based on a similar network architecture, and that provides a suitable approach for myocardial segmentation. Our benchmark shows that the automatic myocardial segmentation takes less than 22 seg. for a volume of 128~x~128~x~13 pixels in a 3.1 GHz intel core i7.