A real-time EIT imaging system based on the split augmented Lagrangian shrinkage algorithm. (November 2017)

• Record Type: Journal Article
• Authors: Zhou, Yongkang; Li, Xiaoqiong
• Is Part Of: Measurement. Volume 110(2017)
• Page Start: 27
• Page End: 42
• Publication Date: 2017-11
• Subjects: Electrical impedance tomography -- Online dynamic display system -- Split augmented Lagrangian shrinkage algorithm; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2017.06.018
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A fast EIT image reconstruction method was proposed based on the SALSA. A high-precision EIT hardware system with 16 electrodes was constructed based on FPGA. The SALSA can significantly reduce computation time while guaranteeing image quality. The imaging speed of final system is 32 fps. Abstract: With advantages of portability, low-cost and noninvasive detection, electrical impedance tomography (EIT) has great potential application value as a bedside monitor. Because image reconstruction in EIT involves addressing serious ill-posed inverse problems, realizing EIT online dynamic display still remains a challenging work. In this paper, an EIT imaging system with the function of online dynamic display was designed. Split augmented Lagrangian shrinkage algorithm (SALSA), which transforms an unconstrained optimization EIT inverse problem into an equivalent constrained optimization problem, was adopted in the proposed system. On the basis of EIT imaging simulation, a high-precise Compact PCI-based EIT system was implemented with FPGA. And experiments of EIT cylinder imaging and in-vivo monitoring were conducted to verify the feasibility of the system. The results indicated that the online dynamic display frame rate of the system, which was centered with a computer (Intel i7 2.2 GHz, 4 GB RAM), was up to 32 fps.