Automatic positioning method and device based on deep learning and radiotherapy equipment

Abstract

The invention discloses an automatic positioning method and device based on deep learning and radiotherapy equipment, and the method comprises the following steps: S1, fixing a patient to be subjected to radiotherapy on a treatment bed, and collecting a DR image of a corresponding part; s2, inputting a CT image collected during planning into the trained U-Net model to obtain a segmentation result of a specified part, and reconstructing to generate a DRR image; s3, respectively inputting the DR image obtained in the S1 and the DRR image obtained in the S2 into a CycleGAN model to obtain a DR image only containing a specified part; s4, registering the DR image only containing the specified part obtained in the step S3 with the DRR image to be registered to obtain a positioning deviation; and S5, controlling the movement of the treatment bed according to the positioning deviation obtained in the step S4, and realizing automatic positioning. According to the method, the quality of the to-be-registered image is effectively improved, the registration precision is improved, the prediction time during positioning is shortened through a deep learning mode, automatic positioning is finally achieved, manual errors are avoided, and the positioning efficiency is improved.

Description

technical field [0001] The invention belongs to the technical field of radiotherapy, and in particular relates to an automatic positioning method, device and radiotherapy equipment based on deep learning. Background technique [0002] At present, tumor radiotherapy is generally carried out in stages, and each treatment will be fixed repeatedly with corresponding thermoplastic films, negative pressure bags and other devices in conjunction with laser light according to the position fixation during positioning CT scanning and the reset situation during simulation. . However, due to various reasons, there is still a certain deviation in the positioning of the patient, and the deviation is mostly between a few millimeters and one centimeter, or even up to several centimeters. [0003] In order to realize image-guided radiotherapy, existing technologies mainly include: selecting ROI (Region of Interest) area according to diagnostic CT and treatment plan, and combining CBCT or oth...

AI Technical Summary

Problems solved by technology

[0004] In particular, MV image registration is the most difficult problem. The image contrast is low and the modal difference is large, which leads to the failure of traditional registration algorithms based on image gray value density.

In addition, iterative registration algorithms based on gray value density are usually very time-consuming and poor user experience

Feature-based registration algorithm, feature selection and extraction are difficult, requiring technicians to do it manually, time-consuming and labor-intensive, and prone to human errors

Images