Scoliosis angle measurement model training method and scoliosis angle measurement method

The invention relates to the technical field of scoliosis angle measurement, in particular to a scoliosis angle measurement model training method and a scoliosis angle measurement method. According to the method, firstly, an image segmentation network is trained to obtain a spine form segmentation network, the spine form segmentation network can predict a spine area, a spine center line and a spine boundary line, and then the spine area, the spine center line and the spine boundary line are input into an angle regression network to obtain a scoliosis angle regression network. In conclusion, the three types of spine form information are adopted, the scoliosis angle regression network can be helped to better learn the position of the spine in the image, the scoliosis angle is accurately output, and in other words, the calculation error of the scoliosis angle can be reduced. 本发明涉及脊柱侧弯角测量技术领域，具体是涉及一种脊柱侧弯角测量模型训练方法和脊柱侧弯角测量方法。本发明首先训练图像分割网络，以得到脊柱形态分割网络，本发明的脊柱形态分割网络能够预测脊柱区域、脊柱中心线和脊柱边界线，之后将脊柱的区域、中心线和边界线这三个形态输入到角回归网络，以得