Lung imaging method based on V-ResNet

Abstract

The invention provides a lung imaging method based on V-ResNet. A V-ResNet network structure is a 51-layer deep neural network model composed of a pre-mapping module, a feature extraction module, a depth imaging module and a residual denoising module which are connected in sequence. According to the network, a boundary voltage signal is mapped into a sequence describing field domain feature distribution by using a pre-mapping module, so that the under-qualitative problem of an EIT inverse problem is solved; a deep stack type V-shaped structure network similar to a coding and decoding structure is utilized to effectively solve nonlinearity and ill-conditionality of an EIT inverse problem. The lung image reconstructed by using the V-ResNet network provided by the invention is clear and accurate in boundary, and the algorithm has good robustness and generalization ability.

Description

technical field [0001] The invention belongs to the field of tomography, and proposes a novel deep network model using an encoding-decoding network structure combined with a residual module for image reconstruction of lung respiratory impedance. Background technique [0002] Electrical tomography is a process tomography technology based on the different electrical properties of the medium in the measured area. The imaging principle is to obtain the spatial distribution information of the medium in the measured area based on the electrode sensitive array, process and transmit the electrical signal as a carrier, and use an appropriate information reconstruction algorithm to reconstruct all the information of the spatial distribution of the medium in the measured area. [0003] The key of electrical tomography technology is to solve the direct problem and the inverse problem, and the direct problem is the basis for solving the inverse problem. The positive problem can be attri...

Images