The invention relates to the technical field of medical
image segmentation, in particular to a hip joint
proximal femur segmentation method based on an improved U-Net neural network. The method comprises the following steps of: 1, acquiring a patient
femur CT image, and preprocessing; 2, constructing a sample
library, and marking a
femur region to obtain a
training set and a
test set; 3, performing data enhancement on the
training set; 4, constructing an improved U-Net neural
network model; 5, training the improved U-Net neural
network model; and 6, testing the trained improved U-Net neural
network model, outputting a segmentation result, and evaluating the model to obtain model performance. According to the hip joint
proximal femur segmentation method, the accuracy of segmentation of the hip joint
proximal femur is improved, the segmentation result can serve as an important reference basis for
femoral prosthesis reconstruction, the accuracy, objectivity and reliability of diagnosis are improved, and the method has important significance on development of target skeleton segmentation in medical images.