Zhu Peifei, Li Zisheng
Hitachi, Ltd., Research and Development Group, Tokyo, Japan.
J Med Imaging (Bellingham). 2018 Oct;5(4):044503. doi: 10.1117/1.JMI.5.4.044503. Epub 2018 Nov 20.
The extraction of six standard planes in 3-D cardiac ultrasound plays an important role in clinical examination to analyze cardiac function. A guideline-based learning method for efficient and accurate standard plane extraction is proposed. A cardiac ultrasound guideline determines appropriate operation steps for clinical examinations. The idea of guideline-based learning is incorporating machine learning approaches into each stage of the guideline. First, Hough forest with hierarchical search is applied for 3-D feature point detection. Second, initial planes are determined using anatomical regularities according to the guideline. Finally, a regression forest integrated with constraints of plane regularities is applied for refining each plane. The proposed method was evaluated on a 3-D cardiac ultrasound dataset and a synthetic dataset. Compared with other plane extraction methods, it demonstrated an improved accuracy with a significantly faster running time of . Furthermore, it showed the proposed method was robust for a range abnormalities and image qualities, which would be seen in clinical practice.
三维心脏超声中六个标准平面的提取在临床检查中对分析心脏功能起着重要作用。本文提出了一种基于指南的高效准确标准平面提取学习方法。心脏超声指南确定了临床检查的适当操作步骤。基于指南的学习理念是将机器学习方法融入指南的每个阶段。首先，应用具有分层搜索的霍夫森林进行三维特征点检测。其次，根据指南利用解剖学规律确定初始平面。最后，应用结合平面规律约束的回归森林对每个平面进行细化。该方法在三维心脏超声数据集和合成数据集上进行了评估。与其他平面提取方法相比，它在运行时间显著更快的情况下，精度有所提高。此外，它表明该方法对于临床实践中会出现的一系列异常情况和图像质量具有鲁棒性。
