Objective It is an urgent need for patients with mechanical ventilation and clinicians to monitor the process of pulmonary ventilation with real-time continuous images at the bedside. Electrical impedance imaging (EIT) of the lung can reflect the distribution of changes in the electrical characteristics of the chest caused by breathing, which has a natural advantage in the monitoring of pulmonary ventilation. The purpose of this paper is to establish a radial basis function neural network (RBFNN) based weighted frequency-difference EIT (wfd-EIT) method to achieve high spatial resolution imaging of pulmonary ventilation.Methods The wfd-EIT method was used to describe the conductivity distribution of the thoracic cavity in real time, and then the target region was visualized and its boundary information was accurately identified by the RBFNN. Firstly, through numerical analysis and simulation, 2 028 simulation samples were established by COMSOL and MATLAB software at each excitation frequency, which were divided into training set and test set to verify the feasibility and effectiveness of the proposed imaging method. Secondly, in order to verify the simulation results, a lung physical model was established. Biological tissues with low conductance characteristics were selected to simulate the ventilation area of the lung, and the imaging experiment was conducted on it. The quantitative data of image correlation coefficient (ICC) and lung region ratio (LRR) were used to measure the accuracy of the imaging method.Results The wfd-EIT method can reconstruct the image at any time and accurately reflect the electrical characteristics distribution of the target region. The algorithm based on RBFNN can enhance the imaging accuracy of the target region with ICC reaching over 0.94, which can better highlight the boundary contour information.Conclusion The wfd-EIT imaging method utilizes the simultaneous measurement of multi-frequency impedance spectra to realize rapid visualization of the target area, and combines the advantages of the RBFNN in approximating arbitrary non-linear functions to achieve accurate identification of the electrical characteristics changes in the target area, which lays theoretical and technical foundations for EIT image monitoring of clinical pulmonary ventilation in the next step.