Which MCU + BT/BLE for heart rate and ECG monitoring (from TI ADS1293)

[victus]

I am working on a portable wireless heart monitor with Bluetooth interface. The idea is to measure ECG signals for at least 1 day straight and then transmit data to the Android device. According to my research I believe that TI's ADS family would probably be the most suitable for my case, specifically ADS1293, which is in short all-in-one ECG chip.

I am now searching for a suitable microcontroller with Bluetooth capability which would acquire the signals from ADS1293, save them to external flash memory and then transfer them to the phone over BT/BLE. The original idea was to work with Arduino, but then I came across SoC TI CC2640. It obviously looks really suitable for my case at first glance since it's dimensions are small and it also has low power consumption, but so does Nordic nRF52832. I've already searched all related forums for comparison between these two and people seem to prefer nRF52832 a bit more over CC2640. However, TI already has a very similar project design like mine here. Code is already provided, but the project is designed with CC2541, which is not compatible with CC2640 though, but I suppose that this would still help with coding.

Since I basically don't have any previous experience with these platforms, I was wondering which MCU+BT/BLE would you recommend to me? An important factor here is also that the development process should be as cheap as possible since I am a student, so the price of debuggers, development boards and software should also be taken into consideration.
Should I rather choose any other MCU e.g. like RFdroid or would you have any other recommendations? I would really prefer a well supported MCU for faster development process and quicker learning curve.

Thanks for help and any suggestions in advance.

[epierre]

@victus said:

. According to my research I believe that TI's ADS family would probably be the most suitable for my case, specifically ADS1293, which is in short all-in-one ECG chip.

just a quick warning, a real ECG has 7 electrodes, saying you want to do it with less must require a serious warning on the accuracy of the results. Also the mesure of some pulse for example PQRST are in millisecond so you need to have the capacity to read and store values so they are meaningfull.

also the HRM which is quite common with now handheld watches and gadgets are mesured on the wrist which is not the best way as it is too close to the end of the member. Those such as polar on the torso have more accuracy.

[sundberg84]

I am a nurse (not active for some years now) with 3 years experience in a CICU - if this is a serious project i might assist you in some ways... let me know.

[victus]

@epierre I have already tried to find comparison between the different number of channels/electrodes in Holter/ECG systems, but I could not fine anything explanatory. Do you perhaps have any literature so I could have a look at it?
Regarding the HRM measurement, I guess it would be the best way to just do it with electrodes since I will need them for ECG measurements anyway?

@sundberg84 of course, any help would be greatly appreciated. Let me write down some more questions and then please let me know, how could you assist.

I would like to kindly ask you to have a look at the TI's project Wireless Heart Monitor with Bluetooth Low Energy. It is supposed to be a fully functional ECG monitor designed with ADS1293 and CC2541. PCB and code are provided as well, so as far as I see it, the fastest solution would be to just use that reference design and modify it in a way that it would store the measurements to external flash (SD) instead of sending it to phone in real-time. In that case I could devote more time for development of the Android app. So one of the drawbacks I see here is that I would need to use IAR Embedded Workbench because of 8051 which is very expensive. I know that there is evaluation version available, do you perhaps know if it is fully functional? Because in that case I suppose I could use e.g. VMware and extend trial period for as long as needed?
Another downside I see when comparing it to CC2640 is the power consumption. Do you think that this would be a problem? Would CC2640 make a big difference?

If I really need to use an alternative to CC2541, then I suppose it should be overall much better MCU to actually be worth it. I suppose that TI CC2640, Nordic nRF51 and Nordic nRF52 would be the best choice therefore? How about BlueGiga e.g. BLE113, I have have seen several good opinions on it? It is quite expensive though since this is only a school project. DKBLE Development Kit costs about 200 $ and I suppose I cannot really do anything without it, right?

[sundberg84]

@victus Im not familiar with this components you mention, you need to test yourself, cant help you with the technical stuff.

It does not seem to be a fully functional ECG but a heart rate monitor. As epierre said above, you need to define if you want to monitor or have a fully functional ECG, its a big difference. With a monitor all you get is pretty much your heartrate and you can detect arytmias. A fully working ECG is normally made with 12 leads and is used to in detail know how the electrical depolarisation from different time and direction/place within the heart muscle works. Holter is a example of a heart rate monitor over time.