Knee joint CT image automatic segmentation method and device and electronic equipment

Abstract

The invention relates to a knee joint CT image automatic segmentation method and apparatus, and an electronic device. The method comprises the steps of obtaining a knee joint original image; processing the original knee joint image according to a preset image processing rule to obtain a target image, wherein the target image comprises a background region and a target region; determining a target area according to a preset area determination rule and the target image, wherein the target area comprises a thighbone area, a tibia area, a fibula area and a patella area; determining a patella region according to a preset first determination rule and the target image; filling the patella region according to a preset filling rule to obtain a patella region boundary line; and determining a femur region boundary line, a tibia region boundary line and a fibula region boundary line according to a preset second determination rule. The knee joint CT image segmentation method has the advantages that the knee joint CT image can be automatically segmented, and the workload of doctors is reduced.

Description

technical field [0001] The present application relates to the field of total knee arthroplasty, and in particular, to a method, device and electronic device for automatic segmentation of knee CT images. Background technique [0002] Knee joint disease is a common clinical disease, which seriously affects people's health and life. Accurate segmentation of the knee joint is an important part of the diagnosis of orthopaedic diseases. Traditional methods of bone and joint segmentation are prone to misdiagnosis and misdiagnosis, thus delaying the treatment of patients and even causing medical malpractice. [0003] At present, for the segmentation of knee CT images, the semi-automatic segmentation method that relies on threshold is mainly used. It does not have high accuracy for the segmentation of fractures and bone fragmentation, and the segmentation process is very time-consuming and requires more participation of doctors; In recent years, with the development of deep learning...

AI Technical Summary

Problems solved by technology

The accurate segmentation of the knee joint is an important link in the diagnosis of orthopedic diseases. The traditional bone joint segmentation method is prone to missed and misdiagnosed situations, which will delay the treatment of patients and even cause medical accidents.

[0003] At present, for the segmentation of CT images of the knee joint, the semi-automatic segmentation method mainly relies on the threshold value, which does not have high accuracy for the segmentation of fractures and bone fragments, and the segmentation process is very time-consuming and requires more participation of doctors; In recent years, with the development of deep learning, segmentation methods based on convolutional neural networks have emerged in an endless stream, but the direction of deep learning requires a large amount of labeled data, which undoubtedly increases the workload of doctors.

Images