Assuming one cleanly modeled a multimodal aftereffect (i.e. visual+audio) I'd expect additional responses in auditory and audiovis integration areas to the inducer stims. Based on this work, I'd hypothesize broader responses to aftereffects with some cortical feedback contributions.

Visual only

The standard dev of resting time series is correlated with CVR probably via voxelwise vessel diameters & baseline CBF (i.e. doi.org/10.1016/j.ne... ). This new method might be more precise, but I'm curious how much better this method matches baseline CBF vs simply the stdev of the resting state TS.

Cortical influences of afterimages may not be essential for day-to-day life, but it's just one example how a behavioral variable we rarely probe may be linked to some of this whole-brain activity. 5/5

As for what the whole brain does, our new preprint looks at afterimages and shows both variations in human influences and a lot of cortical influences bsky.app/profile/dani... 4/5

Localization is not everything but it's only part of the story. One of my core take homes from that work is that much of what we either localize or define as part of networks is limited by our data quality. If collecting 9X data gives diff results, more humility is needed about false negatives. 3/5

Javier's initial whole-brain activity fMRI paper & our followup definitely show whole brain activity to a visual stimulus, but the response is an order of magnitude larger in classically localized visual regions. pubmed.ncbi.nlm.nih.gov/25405938/ 2/5

Brain damage / lesion studies partially answer that some localized regions are essential for behavior and some have more subtle (if any real-time effects). 1/5

Some subject variables such has age or clinical conditions can systemically may vary with dropout: head motion, heavy breathing, head size, metal in mouth...
Suggest calculating voxelwise & ROIwise TNSR (mean/ detrended stdev). If TSNR varies with the behavioral variable, that's a likely cause. 2/2

One possibility is signal dropout. OFC & accumbens are both in areas susceptible to dropout. If an area has significantly less signal or signif more noise, I'd expect all correlations to that area to drop. 1/2

The 4% of survey respondents who agree the Earth is flat, but don't disagree that the Earth makes a yearly orbit around the Sun are either messing with the poll or were really confused by the questions.

"Mind Fixers" by Anne Harrington. Expansive and critical history of biological psychiatry. Uneven at points and I have critiques of some claims, but still a good read.

Currently reading "Until Proven Safe: The history & future of Quarantine" by Nicola Twilley & Geoff Manaugh. Great history that leads to more questions than easy answers.