A Flexible Image-Guided Shape Reconstruction Framework for Electrical Impedance Tomography

Electrical impedance tomography (EIT) owns the advantages of safety, high temporal resolution, and functional imaging characteristics, thus is considered a promising medical/industrial imaging modality. However, due to the ill-posedness of its inverse problem, the spatial resolution of EIT is low, which greatly impeded its practical applications. A flexible image-guided inclusion boundary reconstruction (IGBR) framework for EIT is proposed to alleviate the problem. A statistical inverse problem framework for IGBR is established to directly reconstruct the inclusion boundary and the corresponding conductivity distribution. The morphology prior information is extracted from the images obtained from other high-resolution modes and is used to guide the boundary reconstruction (BR) process in the form of a conditional probability model. A series of simulation and experimental studies are carried out for different application backgrounds. The lung EIT imaging guided by computerized tomography (CT) image and EIT inclusion detection guided by B-ultrasound image is simulated and analyzed, respectively. The results show that the proposed method has high applicability to prior images of different modes. It not only achieves high-precision shape reconstruction but also high-precision conductivity estimation under the guidance of accurate prior images. Meanwhile, the proposed method has high robustness to the possible influence of noise in prior images or incomplete structural information. Even if the information in the prior image is biased or incomplete, the proposed method can still achieve high precision boundary reconstruction and conductivity estimation. The results of quantitative analysis show that compared with the method without image prior, the proposed method has an average reduction of 70% in area error (AE), while the conductivity estimation error is reduced by about five times on average.