AUTHOR=Shoaib Muhammad Ali , Lai Khin Wee , Chuah Joon Huang , Hum Yan Chai , Ali Raza , Dhanalakshmi Samiappan , Wang Huanhuan , Wu Xiang TITLE=Comparative studies of deep learning segmentation models for left ventricle segmentation JOURNAL=Frontiers in Public Health VOLUME=Volume 10 - 2022 YEAR=2022 URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2022.981019 DOI=10.3389/fpubh.2022.981019 ISSN=2296-2565 ABSTRACT=One of the main factors contributing to death across all age groups is cardiovascular disease. Analyzing the left ventricle (LV) from 2D echocardiographic images is a common medical procedure for heart patients. As a result, LV segmentation is a critical task in medical practice. Automated LV segmentation is a pressing need. We designed an automated system that uses echocardiography images to generate LV segmentation using deep learning techniques. The amount of training data required to obtain highly precise segmentation results is also investigated. In this work, three cutting-edge convolutional neural network architectures (SegNet, Fully Convolutional Network, and Mask R-CNN) are designed and implemented to segment the LV. In addition, an echocardiography image dataset is generated, and the amount of training data is gradually increased to measure segmentation performance using evaluation matrices. The pixels accuracy, precision, recall, and dice similarity coefficients (DSC) are applied to evaluate the three models. The Mask R-CNN model outperformed the other two models in these evaluation matrices. As a result, the Mask R-CNN model is used in this study to examine the effect of training data. For 4000 images, the network achieved 98.76 % mean accuracy, 92.21% DSC value, 96.81% recall, and 95.13% precision value. Relatively, the Mask R-CNN outperformed other architectures, and the performance achieves stability after 4000 training images when the model is trained.