Self-adaptive radiotherapy structure automatic sketching method and device, and storage medium

Abstract

The invention belongs to the technical field of radiotherapy, and relates to a self-adaptive radiotherapy structure automatic sketching method and device, and a storage medium. The method comprises the following steps: inputting patient data; preprocessing the medical image of the patient; classifying and positioning a human organ structure in the patient image by using a first-stage deep neural network; segmenting and sketching the human organ structure by using a second-stage deep neural network according to the classification and positioning result; and outputting the segmentation and sketching result. The self-adaptive radiotherapy structure automatic sketching method provided by the invention gives full play to the advantages of full-automatic rapid calculation of image classificationand segmentation by the deep neural network, avoids manual intervention in the classification and segmentation process of the radiotherapy structure, and can complete full-automatic sketching processing according to the image data, the self-adaptive scanning part and the organ structure. The time of dozens of minutes or even hours of traditional sketching work needing manual intervention is shortened to several minutes.

Description

technical field [0001] The invention belongs to the technical field of radiotherapy, and relates to a method, equipment and storage medium for automatically delineating an adaptive radiotherapy structure. Background technique [0002] Tumor is an important disease that endangers human health. With the aging of the social population, its morbidity and mortality will continue to increase. Radiation therapy is one of the important means of tumor treatment. [0003] Tumor radiotherapy endangers organ delineation is the premise to ensure the safety of radiotherapy. Currently, there are two common delineation methods: one is manual delineation, that is, for each organ, the doctor uses the paintbrush or brush in the software according to the CT / MR image content On the image, each organ at risk is manually delineated one by one according to the contour boundary; the efficiency of pure manual delineation is very low, and it often takes several hours to complete the delineation work...

AI Technical Summary

Problems solved by technology

These tools or methods require manual intervention or manual interaction according to the structure of each organ, and the efficiency is somewhat higher than that of pure manual delineation; but it still takes a while to complete. According to the experience of doctors, it usually takes 20 to 30 minutes to complete.

Images