Alexander V Lebedev

Karolinska Institutet | KI · Department of Clinical Neuroscience

Dear Colleagues,

I am about to start working with some aversive learning paradigms in humans and wanted to ask for some suggestions on measuring fear responses to threat/mild pain stimuli. 

Specifically, I wonder if you could share some tricks and tips on how to improve quality and reliability of galvanic skin response (which, I know, is inherently noisy and can be tricky to collect in some subjects) or maybe recommend some good ways/techniques to complement it (e.g. by combining with heart rate variability, etc).

Many thanks in advance!

Dear Colleagues,

I wonder if anyone can recommend a specific pipeline for diffusion MRI data preprocessing with subsequent reconstruction of adjacency matrices. Preferably, with step-by-step instructions and straightforward strategy for scripting/running multiple subjects.

I would also appreciate if you could suggest an optimal tractography algorithm to do this for 70-direction, b=2200 s/mm2 diffusion data.

Thank you very much in advance!

Dear Colleagues,

I have recently found a very nice "Ego Pathology Inventory" based on Karl Jaspers' domains of self-consciousness (Scharfetter, 1981).

Can anyone confirm that the correct scoring strategy is just to sum up all "yes" answers per domain (e.g., 3 "yes" and 2 "no" answers to questions 10-14 imply score 3 in the "Consistency/coherence" domain).

Or is it 1-to-4-point scoring? (i.e. "amost never" [1],  "sometimes" [2], "often" [3] "almost always" [4])

I wonder if anyone has ever experienced running graph theoretical (GT) analysis for task fMRI. If so, what are the general recommendations, precautions, pitfalls?

For instance, given that for this kind of analysis long sessions are strongly recommended, do you suggest to acquire images during a particular task within one large block, or do you think it is still better to use a regular block design, then "cut" the data and reconstruct connectivity matrices for different conditions separately? What would be the recommended block length then?

Dear Colleagues,

I would like to discuss the following questions pertaining to the so-called fMRI data scrubbing (http://www.humanconnectome.org/hosted/docs/Power-et-al-NeuroImage.pdf):

1. How and when do you think it is better to do scrubbing (e.g., Is it better to remove spikes before/after bandpass filtering, before independent component analysis or prior to functional connectivity analysis [for example, with mancovan]?)

2. How many volumes associated with "bad event" do you usually remove? (e.g., only 1 volume, or also neighbouring volumes)

3. How do you threshold your scrubbing? (e.g. FD-threshold=0.5mm)

4. Do you interpolate between your time-points afterwards? If so - how? (Nearest Neighbour, Linear, Cubic Spline)

5. When do you exclude a subject from the analysis completely? (e.g., how many volumes one should miss in %s?)

Any comments and suggestions are very welcomed.

I am currently trying to threshold my SPM t-maps with AlphaSim (REST toolbox implementation) that employs Monte Carlo simulations for the control of types I and II errors.

I have a question pertaining to the values that I should put into viewer after the simulation is done in order to set an appropriate threshold…

My original voxel size is 3.3x3.3x.3.3. Diagonal is about 4.67 - so I set 5 as radius in AlphaSim (Cluster Connectivity criterion).

Smoothness was estimated at 7.

I run 5 000 simulations with individual threshold of 0.05 (using my t-maps and mask.img) and got these values (full report is attached):

Cl Size Frequency Cum Prop p/Voxel Max Freq Alpha

1 24096 0.100189 0.046730 0 1.000000

2 17288 0.172071 0.046606 0 1.000000

3 12540 0.224211 0.046429 0 1.000000

…

1060 0 0.999796 0.000314 0 0.049000

(I Assume this is a cluster size that I should use.)

…

So, now I should set my p-value to 0.05 and select cluster-size of 1060 voxels.. Am I right?