Firmware development and testing of semi-invasive multichannel esophageal ECG recorder for 3D mapping of electrical heart activity

The Multichannel Esophageal ECG recorder was developed as a long-term recording diagnostic device to overcome the limits of normal surface ECGs in the detection of arrhythmias. The surface ECG is a non-invasive device used nowadays by cardiologists as a method of first screening in case of arrhythmias that are difficult to diagnose if not periodical. The limitations of this technology can be caused by irritation of the electrodes on the skin or lack of sensitivity to P waves. Against this background, the device designed by the research team at Institute for Human Centered Engineering HuCE (Bern University of Applied Sciences BFH) has proven to be more sensitive to arrhythmia detection. Subsequently, given the needs of cardiologists to have a device that would allow to obtain a 3D map of the electrical heart activity that does not require invasive methods or the use of ionizing radiation, the attention of the research team shifted to the definition of a new esophageal catheter. With a number of electrodes equal to 64, a innovative design of them, a software for the definition of the signal’s position in 3 dimensions, this seems to be possible, but a new acquisition system is required. During this master thesis work I focused on putting this acquisition system into operation in order to be able to receive and display the correct input signal. For this phase I used, as a test signal, a square wave signal of 2mVpp at frequency of 1 Hz, to have amplitude and frequency comparable to a normal ECG signal. First, a study of the hardware was necessary to understand its operation, using as documentation the datasheets of the components. In this phase using the test signal, the output displayed was not comparable to a square wave. The hypothesis of the failure was attributed to noise of the PCB board being composed of 4 layers and being itself complex. In order to validate this hypothesis, I decided to proceed by dividing the entire system from the acquisition of the signal to the display of the same in blocks. The components have been tested, checking for short circuits and correct voltage values. Then I focused on the firmware where problems related to lack of synchronization between the different threads treated by the microprocessor have been solved. The firmware was written in #C language, using Code Composed Studio and the libraries offered by Texas Instrument. About the visualization software in which the raw data received is also processed, I decided to use only the processing part because the visualization of the signal had a limited number of samples. So once processed, the data is saved in files via the appendance mode.In this way they can be used directly from the new software on which part of the research team is working or simply used offline by me to view the data on my MATLAB script during the tests. Finally, after defining a new prototype of working firmware and software, I focused on the definition of the tests to verify the performance of the correct visualization of the ECG signal.First of all we will test the all channels with the use of a square wave to evaluate if there are problems of connection at the hardware level.Subsequently a first test using surface electrodes for the acquisition of the surface ECG signal will be used to evaluate the correct visualization of this after the filtering operation. The performance of the device has been evaluated and compared with the acquisition system used previously.