A novel left ventricular volumes prediction method based on deep learning network in cardiac MRI

Accurate estimation of left ventricle (LV) volumes plays an essential role in clinical diagnosis of cardiac diseases using MRI. Conventional methods of estimating ventricular volumes depend on the results of manual or automatic segmentation of MRI. However, manual segmentation of MRI sequences is extremely time-consuming and subjective, and automatic segmentation is still a challenging task. Therefore, this study aims to develop a new LV volumes prediction method without segmentation, motivated by deep learning technology and the large scale cardiac MRI (CMR) datasets from the second Annual Data Science Bowl (ADSB) in 2016. The experiments results shows that the predicted LV volumes have high correlation with the ground truth. These results prove that the proposed method has big potential to be researched and applied in clinical diagnosis and screening of cardiac diseases.