Some modifications to the AD5940 Impedance Spectroscopy system are required in order to accomplish an impedance tomography approach. The AD5940 has the required set of filters and amplifiers for impedance data collection, but can only support up to seven inputs while 16 in total are needed for the planned EIT. This can be remedied by routing the leads of the four-electrode setup to different locations on an electrode ring using a custom multiplexer circuit. A multiplexer is a type of digitally-controlled switch that can rapidly change between multiple analog signal inputs and outputs. Our group decided to implement four 16:1 multiplexers so that each of the four electrode leads from the AD5940 could be routed to any of the 16 electrode leads and that the transitions between different positions of the measurement algorithm could be rapid to maximize temporal resolution of the images. For minimal interference with the voltage readings, our group chose the ADG1606 multiplexer, which has relatively small input resistance and leakage current. Control of the multiplexers will be accomplished using the digital pins of the Arduino Mega microcontroller that also controls the AD5940. After designing and testing the circuit, we had a custom PCB turned such that it could interface with the Arduino Mega and AD5940 in a shield configuration.

For the impedance tomography measurement protocol, we selected the adjacent method due to the increased resolution near the edges of the surface. To test the protocol programming, we designed a test in MATLAB interfacing the software with Arduino. The program was designed to set the digital pins of a microcontroller to provide instructions to the multiplexers for each measurement in order. This will route the signal from the correct electrodes on the skin surface. A demonstration of the protocol is given on the left, plotting the position of each electrode over time as the program runs. Moving forward, we plan to implement the program using the SensorPal software suite integrated with the AD5940 to collect measurements.

Currently, we have a MATLAB app demo running for our EIT application utilizing EIDORS. To the left is the current user interface (first image), which will allow the physician to select the leg being measured, and the recorded data set. In the upper right corner of the window is the static reconstruction (in this case an empty saline tank from EIDORS demo data). This will allow the physician to verify the location of the muscle compartments based on anatomical charts (low-conductivity bone structures). The muscle compartments are marked by small circles, from which specific traces will be drawn. The impedance measurements of those specific areas over time will be displayed in the lower left of the UI window. In this case, there is an increase in impedance over time, as the demo data involved moving a low-conductivity object into the lower left of the saline tank (second image). The last section of the UI will be reserved for machine learning, statistical analysis, and ultimately diagnosis.