Implementation of Generalized Back Projection Algorithm in 3-D EIT

Electrical impedance tomography (EIT) is an imaging modality of determining imaging the interior conductivity distribution within an object. In the paper, generalized back projection algorithm (GBPA) is proposed for EIT image reconstruction. The algorithm is based on node back projection algorithm (NBPA), and extends projection range to arbitrary part of the object. To evaluate the proposed method, GBPA is first applied to image reconstruction using a simulated cylinder model. Compared with conventional back projection algorithm (BPA) and NBPA, the images reconstructed by GBPA are more accurate in terms of several metrics used in the paper. In addition, we present image reconstruction results using a 128-channel EIT system to demonstrate the performance of GBPA in real experiments.