Fully-automatic pelvic tumor segmentation method and system, storage medium and terminal

Abstract

The invention provides a full-automatic pelvic tumor segmentation method and system, a storage medium and a terminal thereof, comprising the following steps: preprocessing and data amplification are carried out on images in a meningioma image set to obtain meningioma training data; training U-net convolutional neural network based on the meningioma training data; the pelvic tumor training data were obtained by preprocessing and data amplification of the pelvic tumor images with tumor boundary gold standard drawn in the pelvic tumor image set. Training data of the pelvic tumor based on the training results of the U-net to realize fine-tuning of the full convolution neural network to get the optimal U-net fully convolution neural networks; inputting the acquired pelvic tumor image into the optimal U-Net total convolution neural network to obtain the segmentation results of pelvic tumors. The invention discloses a full-automatic pelvic tumor segmentation method and system, a storage medium and a terminal through migration learning U-net fully convolutional neural network to realize automatic pelvic tumor segmentation, which greatly improves the efficiency of medical interaction and the accuracy of pelvic tumor segmentation.

Description

technical field [0001] The present invention relates to the technical field of medical image processing, in particular to a fully automatic pelvic tumor segmentation method and system, a storage medium and a terminal. Background technique [0002] Among tumors of the musculoskeletal system, there is usually significant heterogeneity in both primary and secondary bone tumors. Especially for complex pelvic tumors, the lesions involve multiple pelvic partitions, and the surgical resection range is large, making the automatic segmentation of pelvic tumors always a big challenge. [0003] At present, tumor boundary image judgment is the link with the lowest efficiency and the most impact on surgical results in medical-industrial interaction engineering. The existing pelvic tumor segmentation methods are usually marked and segmented manually by experienced physicians, which have the following defects: [0004] (1) Time-consuming and labor-intensive, it is difficult to quickly ob...

AI Technical Summary

Problems solved by technology

[0004] (1) Time-consuming and labor-intensive, it is difficult to quickly obtain a large number of segmentation results;

[0

Images