Electrical Impedance Tomography-Based Abdominal Subcutaneous Fat Estimation Method Using Deep Learning

Figure 6

Numerical simulations for testing the proposed reconstruction algorithm with different subcutaneous fat and muscle thicknesses and body shapes. (i) Thick subcutaneous fat. (ii) Thin subcutaneous fat. (iii) Different body shape. (a) True conductivity distributions. (b) Reconstructed images generated by the proposed method. (c) Reconstructed images generated by the regularized least squares method.