Import, process and visualize Sciospec EIT data in the open-source software EIDORS by taking the example of a standard phantom experiment
From research targeted instruments like the EIT16 to fully customized OEM products for bioanalytical, medical and industrial applications Sciospec provides highly specialized solutions for electrical impedance tomography. Flexible channel configurations, frequency sweep modes, scalability up to several hundred channels and broad options for extension through sensor adapters, add-on modules and more make Sciospec EIT systems suitable for a multitude of ambitious applications.
EIDORS (Electrical Impedance Tomography and Diffuse Optical Tomography Reconstruction Software) is a free software for EIT image reconstruction and processing. EIDORS is based on MATLAB, but also works with the free open source alternative Octave and is broadly used and supported by the academic community and EIT experts worldwide. The Sciospec EIT16 – 16 channel EIT device – is a full featured high bandwidth EIT system for ambitious research applications. The EIT data generated using a Sciospec EIT system, like the EIT16, and corresponding Sciospec software can be imported, processed and visualized in the Matlab based free software EIDORS. Sciospec provides appropriate Matlab code examples and guidance to help you getting started with this constellation. This Application Note demonstrates how to import and perform EIT image reconstruction in EIDORS using the Matlab code provided by Sciospec. Some sample data were generated in a standard phantom experiment using the Sciospec EIT16 and corresponding Sciospec software.
Human lung EIT experiment
We reconstructed conductivity changes of human lungs caused by breathing. “The Principal Component Analysis Based Local Reconstruction Method” was applied for lung EIT. Own Matlab GUI was developed which loads .txt data file exported from the Sciospec EIT16 device.
The main idea was to extract lung data from the measured data and to find the corresponding lung region. Low-pass filter and PCA were used which can be considered as machine learning techniques. The standard sensitivity method with Tikhonov regularization was used for reconstruction on the local region with the extracted data
Electrical impedance tomography and new insights
An algorithm for 2D-EIT created in MATLAB® is introduced and described in [2, chap. 5], which is used to improve the quality of the cross-sectional reconstructions and to increase the temporal resolution of the method by one order of magnitude.
The different behavior of an electrical potential in two-dimensional and three-dimensional case is exploited. The idea is first tested and verified by experimental data. The new measuring system “Sciospec EIT 16” from Sciospec Scientific Instruments GmbH is used for measurements. For a possible use in teaching the algorithms-source texts are slim and are available to the public.
The results presented could be used for industry, e.g. in the field of application of extrusion methods for monitoring melting or mixing processes.
New method – Bridge between 2D and 3D
Examination object: 28cm and 10cm diameter pipe water column, anomaly: insulator cylinder of height 28cm and 3cm diameter; Current frequency: 20 kHz; Framerate: 1 / s; Excitation by any two adjacent electrodes; Number of excitation patterns g: 16; Number of measurements per frame: 16 x 16; Reconstruction variable: electrical conductivity in S / m; Notation: u2 {Measurement data at 1 cm water level without anomaly; Corresponds approximately to a 2D problem; U3 {Measurement data at water level of 28 cm without anomaly; Corresponds approximately to a 3D problem; ~ U3: Measurement data at water level of 28 cm with anomaly; T (~ u3): Transformed measured data, input data for the 2D reconstruction algorithm; F2: Solver of the direct problem