BreathUp: Getting closer for the acquisition part


So, Sunday I said that I wanted to improve the server and make some more tests on LoRa and for the beginning of the week, I haven’t really done that…

So what did I do?

First, on Monday I worked with my team mates on a resume of the work that has been done for the project for a presentation we made for our teachers.
Then on Tuesday, I spent a lot of time as a guinea pig to acquire some more ECG and EMG. They are pretty good so we decided to send some of them to E-Respi with a sum up of what has been done. I wrote it with Alexandre and it was quite a long mail to write because writing something with two brains is clearly longer than with one… But we finally sent it and received some encouraging feedbacks. They are real EMG, they are our breathing signals, and they should be clear enough to count the breathing signals and for a doctor to analyze them if necessary.
I’m not saying that they are perfect… We will have huge problems to detect when the patient is moving or speaking and we need to disqualify those moments.
I also spent time on rewriting some Python scripts to filter multiple acquisitions at a time, to export them as csv, to change the acquisition sample time… And they are clearly increasing our efficiency!

Eventually, I did not work so much on the server and the LoRa part, but I might spend more time on it during the end of the week.

See you on Sunday.

Xavier Chapron

BreathUp, I’m alive!

For today the group was supposed to validate the ADC choice for the EMG, and succeed in using the algorithm given by the startup we are working with. Unfortunately it will not be done for tonight.

However like I said in my previous post, step by step we are going forward! With Benoit and with the help of Laurent we had seen that 50Hz noise was there every time, even on a signal of a wave generator (idea of our instructors). At this moment I knew what the main issue was: The only fact to plug our acquisition card on the 50Hz is killing the EMG signal. Even worth, when the computer is no longer working on the battery, the noise is going throw the usb cable. It’s mean none thing: my working time is as long as my battery charge :). (By the way, my hearth is working very well, I am alive!) Then we start to reduce the number of electrodes, it was working quite well, until the moment the battery of my computer was over.

What’s more the algorithm given by I-respi is not as operational as they pretended. Benjamin and Xavier improved it, but not enough. Like Benoit and me, they should finish it by the next two days, and test it on the acquisition we will have done.

The main goal for now it is having the result we hoped for today with the ADC in the next two days. Hopefully I think the next thing we had planned, should not be impacted al lot (like the components choice and the drawing pcb, which is by the way something we had just done in class by Friday)

See you on Wednesday, and I hope I’ll give you great news!