Deep learning-based image segmentation model using an MRI-based convolutional neural network for physiological evaluation of the heart

Abstract

Background and Objective: Cardiovascular disease is a high-fatality health issue in the world. How to accurately measure the cardiovascular function of patients depends on the accurate segmentation of their physiological structure and the accurate evaluation of their functional parameters. Structural segmentation of heart images and calculation of the volume of different ventricular activity cycles are the basis for quantitative analysis of physiological functions and can provide the necessary support for the clinical physiological diagnosis as well as the analysis of various cardiac diseases. Therefore, it is important to explore an efficient heart segmentation algorithm. Methods: A total of 275 nuclear magnetic resonance image (MRI) scans of hearts were collected, analyzed and preprocessed from Huaqiao University Affiliated Strait Hospital, and the data were submitted to our improved deep learning model designed based on U-net network. 80% of the images are the training set meanwhile we use the remaining 20% to be the test set. Based on five time phrases in the stage from end-diastolic (ED) to end-systole (ES), the segmentation findings show that it is possible to achieve improved segmentation accuracy and computational complexity in segmenting the left ventricle (LV), right ventricle (RV), and myocardium (myo). Results: We improved the Dice index of LV to 0.965 and 0.921 in phases ED and ES, and the Hausdorff index decreased to 5.4 and 6.9 in the ED and ES stages, respectively; RV Dice increased to 0.938 and 0.860 in ED and ES, and the Hausdorff index decreased to 11.7 and 12.6 in the ED and ES, respectively; Myo Dice increased to 0.889 and 0.901 in ED and ES, and the Hausdorff index decreased to 8.3 and 9.2 in the ED and ES, respectively. Conclusion: The model obtained in the final experiment can segment the left and right ventricles, along with the myocardium in the cardiac MRI more accurately. This facilitates the prediction of cardiovascular disease in real-time, which has a potential clinical utility prospect.