poetinsf
Senior Member (Voting Rights)
I finally got off my butt and ported my fatigue/PEM prediction model completely over to R. Now I'm trying to figure out how to package it for the users. The best way probably is a website that people can log in with their google/fitbit ID to view their intraday/inter-day data and play with the prediction to see if it is of any use.
I think it will be a worthwhile project even if nothing comes out the project itself. It can at least kickstart an effort to establish objective relationship between exertion and ensuing fatigue/PEM. It would aid the management of ME/CFS for one thing, if you can predict/forecast your PEM/disability. For another, we desperately need an objective way to measure exertion and fatigue for both clinical and research purpose. All we have now is subjective feel or questionnaires which are both tedious and not particularly reliable.
I'd like to eventually turn it into a platform for people to experiment with their own models as well, and crowd-source the effort. Somebody out there with imagination could come up with a better way to predict fatigue/PEM. I still have to figure out ways to provide a secure sandbox for such experimentation. I'll also open-source the software so that people can contribute to improving the software itself.
Meanwhile, I still don't have an automatic way to measure fatigue. I'll eventually look into the fitbit device API and see if there is a way for fitbit to automatically detect ADL and measure fatigue. But that would be a project entirely by itself further down the road. For now, the users will have to enter their fatigue manually. Some people already log their activities, fatigue, supplements, etc. (I, for example, use a google sheet to log everything relevant to ME/CFS.) And it may be possible to pull data from the existing log automatically so that the users don't have to enter them again.
So, the question I have for now is about logging. Do you log your daily activities, symptoms, etc. ? If you do, what tool do you use to log? If not, would you be willing to give it a try if an easy way is available?
I think it will be a worthwhile project even if nothing comes out the project itself. It can at least kickstart an effort to establish objective relationship between exertion and ensuing fatigue/PEM. It would aid the management of ME/CFS for one thing, if you can predict/forecast your PEM/disability. For another, we desperately need an objective way to measure exertion and fatigue for both clinical and research purpose. All we have now is subjective feel or questionnaires which are both tedious and not particularly reliable.
I'd like to eventually turn it into a platform for people to experiment with their own models as well, and crowd-source the effort. Somebody out there with imagination could come up with a better way to predict fatigue/PEM. I still have to figure out ways to provide a secure sandbox for such experimentation. I'll also open-source the software so that people can contribute to improving the software itself.
Meanwhile, I still don't have an automatic way to measure fatigue. I'll eventually look into the fitbit device API and see if there is a way for fitbit to automatically detect ADL and measure fatigue. But that would be a project entirely by itself further down the road. For now, the users will have to enter their fatigue manually. Some people already log their activities, fatigue, supplements, etc. (I, for example, use a google sheet to log everything relevant to ME/CFS.) And it may be possible to pull data from the existing log automatically so that the users don't have to enter them again.
So, the question I have for now is about logging. Do you log your daily activities, symptoms, etc. ? If you do, what tool do you use to log? If not, would you be willing to give it a try if an easy way is available?