The next-generation acquisition system.
This page outlines the beta-test of our new PCIe based acquisition system. As far as we know this is the most advanced neuroscience data acquisition system currently available, especially for freely behaving experiments:
One interfaces for multiple probes
PCIe interface for very low latencies (<100us round-trip possible ).
Runs in Bonsai for easy synchronization of multiple data sources and integration into dynamic/reactive/closed-loop experimental setups.
Communicates with headstages via ultra-thin 0.38mm diameter coax.
Standardization of hardware, firmware, protocols, and APIs make it easy to extend to other data sources or actuators, or to integrate with new software.
Precise 3D tracking with steamVR lighthouse base stations.
IMU with heading and acceleration measurements at 100Hz.
On-headstage LED or laser diode driver for optogenetics.
On-headstage precision bipolar current source for neural stimulation.
Torque-independent commutator: The headstages measure their own rotation, and a motorized commutator keeps the tether tangle-free without requiring any torque measurements. Because everything on the headstage operates via the coax, this simple commutator is all that is needed even for opto/e-stim experiments.
The thin tether together with the commutator make this system very low-impact in terms of natural mouse behavior.
This system was developed by Jon Newman and Jack Zhang (hardware, firmware, and API), Jakob Voigts (hardware), with support, and in coordination with, Filipe Carvhalo, Daniel Aharoni, Josh Siegle, Gonacalo Lopes, and Aaron Cuevas Lopez.
In a nutshell, this is a modular data acquisition interface for neuroscience. It currently supports next-gen open ephys headstages, and will soon support Miniscopes and Neuropixels probes.
The basic system package consists of:
Mouse headstage ($5000)
PCIe interface card and analog&digital io board ($20000)
2x Ultra-thin coax tether ($500)
These prices reflect initial development and production costs, and fact that we need to acquire some expensive IP before distributing the system, and will drop to similar, or slightly higher prices as the current system eventually.
(1) Freely behaving mouse ephys with tetrodes (working now):
The recently developed next-generation Open Ephys drive with 16 tetrodes was designed in tandem with the (64 channel) mouse headstage, the entire package comes in at 3.5g, as implanted with all glue, ground screw etc. Together with the 3d tracking, and torque-free commutator this is now the best tethered mouse recording system we know of. A version for larger animals with 256 channels can be made available if there is interest (link).
The mouse drive implant (available both for conventional 64-channel intan headstages and for the next-gen system), next to the 64 tetrode rat drive.
The mouse drive is well tested and is available now, together with tools required for assembly, and with most of the nonstandard consumables.
The rat implant has been tested in the Wilson lab at MIT, but is not currently available in assembled form. Get in touch if you are interested in this drive and headstage and we can figure out a timeline for making it available.
(2) UCLA miniscope (expected by the end of 2019):
The PCIe system is compatible with a upcoming version of the v4 UCLA miniscope. The feature set of the microscope will remain the same as before but now includes all the features of the mouse ephys headstage (commutator support, 3D tracking, LED driver etc.)
(3) Neuropixels (expected by the end of 2019):
A variant of our headstage plugs into the Neuropixels probe. Our system can now record Neuropixels data, but the configuration of the probe to record from other channels than the default set near the tip of the probe is not functional yet. We expect this to be resolved with the imminent release of the development kit frrom imec which fully documents the probe.