Open Access

Parallel, multi frequency EIT measurement, suitable for recording impedance changes during epilepsy


Cite

Fig. 1

A - Current source implementation. Sine wave generation performed by Arduino Nano with output filtering. Howland current pump used for V-I conversion. B – Example connection of multiple current sources to phantom.
A - Current source implementation. Sine wave generation performed by Arduino Nano with output filtering. Howland current pump used for V-I conversion. B – Example connection of multiple current sources to phantom.

Fig. 2

Output impedance of current source. Measured using HP43284A Impedance Analyser.
Output impedance of current source. Measured using HP43284A Impedance Analyser.

Fig. 3

Stability of current source over time, measured on a resistor phantom. The slight drift over time is thought be caused by device heating.
Stability of current source over time, measured on a resistor phantom. The slight drift over time is thought be caused by device heating.

Fig. 4

Results of imaging test. Blue areas represent an increase in impedance. The first column indicates the position of the perturbation in the tank. The remaining columns show the reconstructed images for each frequency group used. Colour bar indicates percentage change in impedance.
Results of imaging test. Blue areas represent an increase in impedance. The first column indicates the position of the perturbation in the tank. The remaining columns show the reconstructed images for each frequency group used. Colour bar indicates percentage change in impedance.

Fig 5

Results of coin drop experiment. 10 frames are shown, taken 10 ms apart, for each coin dropped. Each frame uses 1 ms of recorded data. The black circle at 0 ms indicates the position at which the coin was dropped.
Results of coin drop experiment. 10 frames are shown, taken 10 ms apart, for each coin dropped. Each frame uses 1 ms of recorded data. The black circle at 0 ms indicates the position at which the coin was dropped.

Fig. 6

(A) Experimental Setup. 60 electrodes were placed on either side of the brain. Seizures were triggered using a 4-AP Model. (B) Data extraction. By low pass (EEG) and bandpass (EIT) filtering, the data was split into the EEG component and the EIT component, from which the impedance change was extracted.
(A) Experimental Setup. 60 electrodes were placed on either side of the brain. Seizures were triggered using a 4-AP Model. (B) Data extraction. By low pass (EEG) and bandpass (EIT) filtering, the data was split into the EEG component and the EIT component, from which the impedance change was extracted.

Fig. 7

dZ values recorded at 2.2 kHz and 2.6 kHz during a single seizure. Seizure onset is at t = 0 . Right and left refers to the site of electrode placement.
dZ values recorded at 2.2 kHz and 2.6 kHz during a single seizure. Seizure onset is at t = 0 . Right and left refers to the site of electrode placement.

Fig. 8

Example of raw recorded data, recorded from the right hemisphere. Solid colours represent the EEG data. dZ values, from 2.2 kHz injection, are plotted on top of each channel in grey.
Example of raw recorded data, recorded from the right hemisphere. Solid colours represent the EEG data. dZ values, from 2.2 kHz injection, are plotted on top of each channel in grey.

List of frequency groups

Group A1800 Hz1900 Hz2000 Hz2100 Hz
Group B1980 Hz1990 Hz2000 Hz2010 Hz
Group C1998 Hz1999 Hz2000 Hz2001 Hz