Bioimpedance Parameter Estimation using Fast Spectral Measurements and Regularizaton

Roberto G Ramírez-Chavarría, Gustavo Quintana-Carapia, Matias I. Müller, Robert Mattila, Daniel Matatagui, Celia Sánchez-Pérez

Research output: Chapter in Book/Report/Conference proceedingConference paperResearch

Abstract

This work proposes an alternative framework for parametric bioimpedance estimation as a powerful tool to characterize biological media. We model the bioimpedance as an electrical network of parallel RC circuits, and transform the frequency-domain estimation problem into a time constant domain estimation problem by means of the distribution of relaxation times (DRT) method. The fredholm integral equation of the first kind is employed to pose the problem in a regularized least squares (RLS) form. We validate the proposed methodology by numeral simulations for a synthetic biological electrical circuit, by using a multisine signal in the frequency range of 1kHz to 853kHz and considering an error in the-art techniques for spectral bioimpedance analysis. We also illustrate its potentiality in terms of accurate spectral measurements and precise data interpretation, for further usage in biological applications.
Original languageEnglish
Title of host publication18th IFAC Symposium on System Identification (SYSID 2018)
PublisherIFAC - PapersOnLine
Pages521-526
Volume51
Edition15
ISBN (Print)2405-8963
Publication statusPublished - 2018
Event18th IFAC Symposium on System Identification: SYSID 2018 - AlbaNova University Center, Stockholm, Sweden
Duration: 9 Jul 201811 Jul 2018
https://www.kth.se/en/eecs/om-oss/konferenser-och-event/sysid2018

Conference

Conference18th IFAC Symposium on System Identification
CountrySweden
CityStockholm
Period9/07/1811/07/18
Internet address

Keywords

  • Biomedical Systems
  • Frequency Measurements
  • Impedance Spectroscopy
  • Parameter Estimation
  • Regression Algorithm
  • Spectrum Analysis

Fingerprint Dive into the research topics of 'Bioimpedance Parameter Estimation using Fast Spectral Measurements and Regularizaton'. Together they form a unique fingerprint.

Cite this