Quantitative Assessment of Image Reconstruction Algorithms in Electrical Impedance Tomography

Electrical Impedance Tomography (EIT) is a noninvasive imaging technique used to estimate the internal conductivity distribution of a region that is either unknown or inaccessible. This is achieved by applying electrical currents to the region and measuring the resulting boundary voltages. The forward problem in EIT is typically solved using the Finite Element Method (FEM), and regularization techniques are employed to stabilize the ill-posed inverse problem during image reconstruction. This study evaluated the performance of two widely used image reconstruction algorithms: the delta conductivity method and the Jacobian (JAC)-based method. Both algorithms were tested on seven phantom images with varying levels of complexity to assess their effectiveness in different scenarios. The average Peak Signal-to-Noise Ratio (PSNR) and Mean Structural Similarity Index (MSSIM) were 35.71 dB and 0.93, respectively, indicating high reconstruction quality. However, the complexity of the images, such as intricate textures or multiple inclusions, resulted in reduced reconstruction accuracy. Although, both the delta conductivity and JAC methods proved effective in EIT image reconstruction, the JAC method shows superior performance in more challenging cases.