search

ucl-pond / pySuStaIn

Subtype and Stage Inference (SuStaIn) algorithm with an example using simulated data.
MIT License
135 stars 64 forks source link

Data Preparation Pipeline/Code #34

Closed lmz123321 closed 2 years ago

lmz123321 commented 2 years ago

Hi, pySuStaIn is a great work, thanks for your effort!

I hope to use SuStaIn to subtype patients from our private Alzheimer's Disease structure MRI dataset.

However, I find it hard to implement the data preparation part (i.e. how to obtain z-scores from the raw MRI images).

Could you share your AD MRI data preparation code? or maybe provide some more detailed pipeline instructions of it?

(I checked the instruction provided in your Nat.Comm. paper, but found it too coarse to reproduce.)

Grateful for your help.

noxtoby commented 2 years ago
  1. Extract biomarkers from your raw MRI, e.g. regional volumes of GM tissue.
  2. Identify your control group => calculate mean of controls m and standard deviation of controls s for each biomarker
  3. For each biomarker x, calculate z = (x - m)/s
  4. Make sure that z increases with progression => swap sign if necessary
lmz123321 commented 2 years ago

Hi, sorry to bother again. @noxtoby

I notice, in ADNI, the control group is defined as those CN subjects with CSF Abeta 1-42 > 192 pg/ml.

I do not understand why we need the restriction of CSF Abeta 1-42 > 192 pg/ml ?

Since we currently do not have the CSF data, can we just use all CN subjects as our control groups?

Thx.

noxtoby commented 2 years ago

Hope this doesn't sound too brutal, but you're asking about a study design choice, not pySuStaIn. This is not the place for that.

For the record, amyloid pathology is a hallmark of Alzheimer's disease and so CN individuals with low Abeta in CSF have high Abeta in the brain and therefore may develop Alzheimer's disease.

lmz123321 commented 2 years ago

My apologies.

I know this is not an appropriate place to ask problems about the SuStaIn paper. I asked here only because I failed to receive any respondence from Dr. Young (both [at]ucl.ac.uk and [at]kcl.ac.uk) for about a weak. Any suggestions?

noxtoby commented 2 years ago

Again, sorry if this sounds harsh, but yours is a basic experimental design question. It's a conversation for your team.

We have provided a valuable tool for you to use. We are not collaborating with you on a research project.

We are busy.