The treadmill is one of the key parts of the setup. It allows the mouse to run freely while it stays head-fixed in the same location with the virtual reality moving around it. To achieve this, it has to satisfy a few key requirements:
- run freely in any direction
- track the ball movement to update the location in the virtual reality
- appropriately sized ball for the mouse so that normal movement is possible
To achieve free movement into any direction, a sphere of 20cm diameter was placed in a cup which has an inner diameter of just over 20cm. At the south pole of the base is a whole through which pressurised air is pumped into the clearance between ball and base.The clearance and quality of fit between cup and bowl dictates how much volume of air per unit time is required.
Different techniques have been used in the past to get a good cup for a 20cm or 8in ball. Harvey et. al. (2009) used an epoxy cast, Niell and Stryker (2010) used a steel ladle and UCL carved it out of a solid block of gold (or so I can only assume given the pricing), but there are numerous other solutions. We used a 250mm polystyrene hollow half sphere with a wall thickness of 25mm which resulted in a very good fit.
We worked the hollow polystyrene half sphere with a hotwire to reduce wall thickness and height. You can find more about that here. The base is not an entire half sphere since that has two disadvantages:
1) the ball is more likely to touch the walls, thus making movement of the ball less smooth
2) excrement of the mice would fall into the clearance between base and ball, potentially jamming the treadmill and certainly painting a nice history of the mouse's recent movement on the ball.
Our current base therefore covers about 2/3 of the bottom hemisphere of the ball. This is enough to produce a robust air cushion for the ball to run on.
The pressurised air supply is regulated with a simple shut-off valve. Not the best solution but it certainly works for the crude pressure regulation we require. Should we require finer control of airflow we can either introduce a valve with fine control or a simple pinch valve after the shut-off valve, at the moment it doesn't look like this will be necessary though.
A 16mm OD (outer diameter) tube runs from the valve to the treadmill. The valve produces a lot of noise due to the high flow speed of the air running through it. To dampen that noise a 20m coil of 16mm OD tubing is rolled up between treadmill and valve. Directly underneath the treadmill is an L-joint because the tube isn't flexible enough to make a sharp 90 degree corner.
Below are a few photos, you can find details about how it was built in the blog section.
|L-joint under treadmill. The base itself sits on 3 normal poles. These provide enough stability for our application.|
|At the bottom you can see the shut-off valve. The blue tube (8mm OD) is the airsupply from the room. After the valve the tube diameter is increased to 16mm OD. The very unprofessionally mounted 20m roll of 16mm OD tube reduces noise from the valve.|
|Small bridges fix the tube to the table before it reaches the treadmill.|
|Here and in the first picture you see the L-joint that directs the pressurised air into the base.|
I will add a video of the running treadmill soon.
Harvey, C. D., Collman, F., Dombeck, D. A., & Tank, D. W. (2009). Intracellular dynamics of hippocampal place cells during virtual navigation. Nature, 461(7266), 941-946. Nature Publishing Group. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/19829374
Niell, C. M., & Stryker, M. P. (2010). Modulation of Visual Responses by Behavioral State in Mouse Visual Cortex. Neuron, 65(4), 472-479. Elsevier Ltd. Retrieved from http://dx.doi.org/10.1016/j.neuron.2010.01.033