-
New Forum Software Installed (Still a few issues but reopening the forum) More details.
S4ME Q&A with Prof Chris Ponting - Question Collection Thread
- Thread starter Andy
- Start date
- Status
Esther12
Senior Member (Voting Rights)
I'm after some insider gossip about the Science Media Centre, but don't know how to phrase a question on that without it sounding like a trap.
Why did the CMRC choose to part ways with the SMC?
What do you think of the 'fact-sheet' they put out about CFS, or they way they profess to speak for the 'scientific-community'?: https://web.archive.org/web/2018032...re.org/cfsme-the-illness-and-the-controversy/
Something like that.
Why did the CMRC choose to part ways with the SMC?
What do you think of the 'fact-sheet' they put out about CFS, or they way they profess to speak for the 'scientific-community'?: https://web.archive.org/web/2018032...re.org/cfsme-the-illness-and-the-controversy/
Something like that.
(Sorry a long complex question)
I have a question around collecting, testing and interpreting samples.
ME becomes worse with exertion and this can last multiple days. Should we therefore consider the body as a dynamic system when taking samples and thus always take them with an understanding around recent levels of exertion.
My concern is that if certain measurements vary in ME not due to the underlying cause but due to a mechanism that is a function of the cause, time and exertion levels over a past time period then without understanding if there is a relationship between recent past exertion levels and given readings we may miss something important that would help understand the mechanism that maps from exertion into fatigue or symptom increase. Equally if there is a relationship between readings and recent exertion then this could lead to random looking readings when sampling a population whilst we don't have all the necessary underlying variables (exertion).
Hence when sampling a population should we have recordings of activity prior to sampling.
Or should we be doing longitudinal studies on a few patients where over time we look at readings against exertion levels and reported fatigue levels. This may help understand if there are relationships without the complexities of the natural variation of measures within a population.
I'm aware that data processing techniques for such sequences may not be well developed although we could look to economic modelling techniques. Also as machine learning has developed in recent years there is some very interesting work in recurrent neural networks that learn longer term predictions (hence demonstrate that relationships are modelled). But such techniques would require massive amounts of data.
So my basic question is how do we take account of the dynamic aspects of ME as a disease where exertion could lead to measurement changes and should detailed studies be done to understand this prior to sampling larger populations of patients?
I have a question around collecting, testing and interpreting samples.
ME becomes worse with exertion and this can last multiple days. Should we therefore consider the body as a dynamic system when taking samples and thus always take them with an understanding around recent levels of exertion.
My concern is that if certain measurements vary in ME not due to the underlying cause but due to a mechanism that is a function of the cause, time and exertion levels over a past time period then without understanding if there is a relationship between recent past exertion levels and given readings we may miss something important that would help understand the mechanism that maps from exertion into fatigue or symptom increase. Equally if there is a relationship between readings and recent exertion then this could lead to random looking readings when sampling a population whilst we don't have all the necessary underlying variables (exertion).
Hence when sampling a population should we have recordings of activity prior to sampling.
Or should we be doing longitudinal studies on a few patients where over time we look at readings against exertion levels and reported fatigue levels. This may help understand if there are relationships without the complexities of the natural variation of measures within a population.
I'm aware that data processing techniques for such sequences may not be well developed although we could look to economic modelling techniques. Also as machine learning has developed in recent years there is some very interesting work in recurrent neural networks that learn longer term predictions (hence demonstrate that relationships are modelled). But such techniques would require massive amounts of data.
So my basic question is how do we take account of the dynamic aspects of ME as a disease where exertion could lead to measurement changes and should detailed studies be done to understand this prior to sampling larger populations of patients?
- Status