A tumor detection method and device based on fusion of bm3d and dense convolutional network

Abstract

The invention provides a tumor detection method and device based on the fusion of BM3D and dense convolutional network, marking similar blocks; randomly discarding and fast marking; optimizing and training the DenseNet network; using the spatial information, structural information and extracted feature information of the input image Reconstruction based on deep training data. Abstract the input spatial information into one dimension to reduce the phenomenon of irreversible initial feature loss. Construct a dense convolutional network that integrates BM3D, use a scalable exponential linear unit activation function instead of a linear non-saturated unit activation function to activate the network, and introduce negative partial parameters to improve network optimization and enhance network robustness, and in each dense block Finally, a maximum pooling layer is added to abstract the image features and extract the core information points of the tumor. At the end of the network, the BM3D aggregation method is used for feature reconstruction, and gradient and spatial information are integrated to improve the network effect. Effectively improves the accuracy of tumor detection.

Description

technical field [0001] The present invention relates to the technical field of medical image processing, in particular to a tumor detection method and device based on fusion of BM3D and dense convolutional network. Background technique [0002] Medical images come from imaging techniques, including computed tomography, magnetic resonance imaging, ultrasound, positron emission tomography, medical ultrasound examination, etc. Medical imaging technology can obtain two-dimensional or three-dimensional images of the corresponding positions of the human body. In a two-dimensional image, the smallest unit element representing specific information is called a pixel; in a three-dimensional image it is called a voxel. Under certain conditions, three-dimensional images can be represented as a series of two-dimensional images, which greatly reduces computational complexity and memory requirements. However, although the medical imaging technology has become more and more mature, the re...

AI Technical Summary

Problems solved by technology

[0005] However, when the traditional CNN or FCN transmits information, there will be more or less problems such as information loss and loss. After the input information or gradient information passes through multiple layers, it is likely to cause gradient disappearance or gradient explosion.

Images