Images in Microns Dataset Stimuli

Hello, I am a student researcher exploring the microns dataset. I was looking into the stimuli that were presented to the mouse. I am looking for static images, but all I can find is videos. Would it not make more sense to present a static image to the mouse for a short period and record the response? This would give us a direct correlation between the stimulus and the response. In case of a video, it is not possible to form this sort of relationship as we are not sure which image/object in the video caused the spike, as the spike could be delayed, or it could well be a result of a complex (compound) input? Please correct me if I am wrong, but this is what I have understood so far.

Hi, thanks for your suggestion. Unfortunately, the experiment is now over and so changes to the visual stimulus are no longer possible. There are advantages and disadvantages of almost every stimulus design you can imagine. We agree that having static images would be useful, but there was limited time to record all the neurons and present all the stimulus and keep the project on schedule. The current stimulus was designed to maximize the chance that the team at Baylor could fit complex dynamic models to predict the response of individual neurons to any stimulus.

We hope to release a dataset in the next year or so that will also contain visual stimuli that are more closely matched the the stimuli used in the Visual Coding datasets, which includes some static pictures, along with more conventional parametric stimuli. This stimulus set will be less rich that the dynamic clips, but easier to analyze as you were suggesting.

Finally as a philosophical aside, I think the problem of determining “what caused the spike” is a complex one, even in the context of static images. Conventional neural coding analysis has used simple reverse correlation techniques to try to estimate linear receptive fields for individual neurons, but we know that these don’t fully describe the visual system and there are a myriad set of non-visual factors that influence the spiking of individual neurons.