Skip to main content
. 2017 Dec 5;12(12):e0188993. doi: 10.1371/journal.pone.0188993

Fig 1. Comparison between existing methods and the proposed method.

Fig 1

(a) The general EIT reconstruction method, (b) an ANN used as an EIT inverse solver, highly sensitive to modeling errors and noise, and (c) the proposed reconstruction method, less sensitive to measurement errors. In (c), the input of the ANN is the output of the inverse solver.