Image recognition method, system and device for liver and gall vessels and calculus and storage medium

Abstract

The invention discloses a image recognition system for liver and gall duct and calculus, which comprises an input module used for acquiring an image to be detected and inputting the image to be detected into a trained sparse convolutional neural network; a sparse convolutional neural network used for carrying out image recognition on the to-be-detected image and segmenting position contours of theliver and gall vessels and gall stones in the to-be-detected image in a recognition result; wherein each convolution layer in a first encoder in the sparse convolutional neural network adopts a smallconvolution kernel, and each convolution layer in other encoders and decoders adopts a sparse convolution kernel; in the first encoding process, the feature maps under different scales are output toall subsequent encoders and decoders, and fusion operation is carried out on the feature maps under the corresponding resolutions. By applying the scheme provided by the invention, image recognition can be more effectively carried out on the liver and gall ducts and stones so as to assist doctors in treatment. The invention further provides an image recognition system and device for the liver andgall ducts and stones and a storage medium which have corresponding effects.

Description

technical field [0001] The invention relates to the technical field of medical image segmentation, in particular to an image recognition method, system, device and storage medium for hepatic bile ducts and stones. Background technique [0002] At present, the surgical treatment of hepatolithiasis is facing many difficulties and challenges, including difficulty in removing hepatolithiasis, unclean stone removal, and easy recurrence after operation. The general plan of stone removal surgery is to perform enhanced CT scan before the operation to obtain two-dimensional image information, and then abstract it into a 3D model in the doctor's mind based on experience. The requirements for the doctor's clinical experience and professional knowledge are very high, and deviations may also occur. In order to make better use of digital imaging technology to assist physicians, medical image segmentation technologies specifically targeting specific organ regions and corresponding lesion a...

AI Technical Summary

Problems solved by technology

For highly deformable stones, as well as bile ducts with small pixel areas and variable shapes, the effect is not particularly ideal

Images