Open Access

Bioimpedance and NIR for non-invasive assessment of blood glucose

Jan-Hugo Andersen
Jan-Hugo Andersen
, 
Olav Bjerke
Olav Bjerke
, 
Fatos Blakaj
Fatos Blakaj
, 
Vilde Moe Flugsrud
Vilde Moe Flugsrud
, 
Fredrik Alstad Jacobsen
Fredrik Alstad Jacobsen
, 
Marius Jonsson
Marius Jonsson
, 
Eirik Nobuki Kosaka
Eirik Nobuki Kosaka
, 
Petter André Langstrand
Petter André Langstrand
, 
Øyvind Grannes Martinsen
Øyvind Grannes Martinsen
, 
Alexander Stene Moen
Alexander Stene Moen
, 
Emily Qing Zang Moen
Emily Qing Zang Moen
, 
Øyvind Knutsen Nystad
Øyvind Knutsen Nystad
, 
Eline Olesen
Eline Olesen
, 
Mahum Qureshi
Mahum Qureshi
, 
Victor Jose Østrem Risopatron
Victor Jose Østrem Risopatron
, 
Simen Kristoffer Ruud
Simen Kristoffer Ruud
, 
Nikolai Stensø
Nikolai Stensø
, 
Fredrik Lindseth Winje
Fredrik Lindseth Winje
, 
Eirik Vetle Winness
Eirik Vetle Winness
, 
Sisay Abie
Sisay Abie
, 
Vegard Munkeby Joten
Vegard Munkeby Joten
, 
Christian Tronstad
Christian Tronstad
, 
Ole Elvebakk
Ole Elvebakk
 and 
Ørjan Grøttem Martinsen
Ørjan Grøttem Martinsen
   | Dec 31, 2019
Journal of Electrical Bioimpedance's Cover Image
About this article
Previous Article
Next Article
Cite
Published Online: Dec 31, 2019
Page range: 133 - 138
Received: Oct 12, 2019
DOI: https://doi.org/10.2478/joeb-2019-0019
Keywords
© 2019 Jan-Hugo Andersen et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Fig.1

Unprocessed NIR spectra for glucose levels below 5.0 mM (blue lines) or above 7.0 mM (red lines).
Unprocessed NIR spectra for glucose levels below 5.0 mM (blue lines) or above 7.0 mM (red lines).

Fig.2

Measured resistance as a function of frequency for glucose levels below 5.0 mM (blue lines) or above 7.0 mM (red lines).
Measured resistance as a function of frequency for glucose levels below 5.0 mM (blue lines) or above 7.0 mM (red lines).

Fig.3

Measured reactance as a function of frequency for glucose levels below 5.0 mM (blue lines) or above 7.0 mM (red lines).
Measured reactance as a function of frequency for glucose levels below 5.0 mM (blue lines) or above 7.0 mM (red lines).

Fig.4

Correlation between measured resistance at different frequencies, and blood glucose level.
Correlation between measured resistance at different frequencies, and blood glucose level.

Fig.5

Correlation between measured reactance at different frequencies, and blood glucose level.
Correlation between measured reactance at different frequencies, and blood glucose level.

Fig.6

Correlation between measured phase angle at different frequencies, and blood glucose level.
Correlation between measured phase angle at different frequencies, and blood glucose level.

Fig.7

Correlation between measured conductance at different frequencies, and blood glucose level.
Correlation between measured conductance at different frequencies, and blood glucose level.

Fig.8

Correlation between measured susceptance at different frequencies, and blood glucose level.
Correlation between measured susceptance at different frequencies, and blood glucose level.

Fig.9

PLS model with three components based on NIR, resistance and reactance. Individually trained until the start of intake of sugar-containing soft drink for the second time (blue dots) and thereafter used for prediction (orange dots).
PLS model with three components based on NIR, resistance and reactance. Individually trained until the start of intake of sugar-containing soft drink for the second time (blue dots) and thereafter used for prediction (orange dots).

Fig.10

PLS model with five components based on NIR, resistance and reactance. Globally trained until the start of intake of sugar-containing soft drink for the second time (blue dots) and thereafter used for prediction orange dots).
PLS model with five components based on NIR, resistance and reactance. Globally trained until the start of intake of sugar-containing soft drink for the second time (blue dots) and thereafter used for prediction orange dots).

Fig.11

ANN model with one hidden layer and five nodes based on NIR, resistance and reactance. Globally trained until the start of intake of sugar-containing soft drink for the second time (blue dots) and thereafter used for prediction orange dots).
ANN model with one hidden layer and five nodes based on NIR, resistance and reactance. Globally trained until the start of intake of sugar-containing soft drink for the second time (blue dots) and thereafter used for prediction orange dots).
eISSN:
1891-5469
Language:
English
Publication timeframe:
Volume Open
Journal Subjects:
Engineering, Bioengineering and Biomedical Engineering, Biomedical Electronics, Life Sciences, Biophysics, Medicine, Biomedical Engineering, Physics, Spectroscopy and Metrology
Journal RSS Feed