Simulation of Neural Network-Multicriterion Optimization Image Reconstruction Technique (NN-MOIRT) for imaging using a 32-channel Brain ECVT sensor. (January 2019)

• Record Type: Journal Article
• Authors: Handayani, N; Hidayanti, K F; Baidillah, M R; Arif, I; Khotimah, S N; Haryanto, F; Taruno, W P
• Is Part Of: Journal of physics. Volume 1127(2018)
• Publication Date: 2019-01
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1127/1/012008
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: NN-MOIRT has been proposed earlier as an alternative for existing image reconstruction algorithms that can accurately show a volumetric image in a cylindrical sensor vessel. The Brain ECVT sensor has a different shape and dimensions compared with a common ECVT sensor, which has a cylindrical shape. Using a different sensor changes the image reconstruction algorithm parameters. Thus, the image reconstruction algorithm should be modified to be able to properly make a reconstruction. In this study, a simulation of the reconstruction of an image from a 32-channel Brain ECVT sensor using NN-MOIRT was conducted. The simulation was performed by varying the position and number of objects in the helmet-shaped Brain ECVT sensor. The alpha parameter (penalty factor) was varied from 10 to 150 with the number of iterations from 1 to 200. The RMSE (root mean square error) was calculated based on the difference between the permittivity distribution of the objects and the reconstructed image. It was found that the NN-MOIRT algorithm is more convergent and more stable for image reconstruction than the ILBP algorithm.