Figures and data

Overview of the Pinpoint interface. (a) Vertical panels show the channel map for two Neuropixels 1.0 probes. Channels are colored by the mouse Common Coordinate Framework region they are inserted in. Area acronyms are shown on the right. (b) The 3D scene shows a transparent view of the mouse brain and models of the Neuropixels probes. On the top right of the user interface (UI), an orientation widget (red, yellow, and blue crosshairs) helps users track the orientation and snap the view to axial, coronal, or sagittal views. The right UI panels include an in-plane slice of the reference atlas, individual area search, stereotaxic coordinates, and at the bottom right a “Share” button which creates a permanent URL to re-load the same scene in another browser. (c) Searching for regions highlights them in the 3D scene as opaque 3D models (highlighted here are parts of visual cortex, hippocampus, and thalamus). (d) Examples of the probes available in Pinpoint. (e) The stereotaxic coordinates relative to a reference coordinate (Bregma, by default) for performing the insertion. The angles (yaw, pitch, and roll) are used to set up the manipulator prior to an experiment. (f) A craniotomy placement tool, combined with a model of the mouse skull, helps users plan surgeries. (g) Example of 3D models that can be placed in the scene, here a “skull cap” used for an implant surgery is shown over the mouse skull, followed by additional examples of experimental hardware including headbars and an imaging lens. (h) Pinpoint detects collisions, to ensure that multi-probe insertion plans are4possible. Colliding models are highlighted in red.

Still frame from the Plan an insertion in Pinpoint video, a six-minute demonstration of key features.

Examples of mouse brain reference atlases included in Pinpoint. (a) Pinpoint supports both histological and anatomically accurate reference atlases. The default axes of the Unity world space represent a standard cartesian coordinate system. To define the space of a reference atlas, such as the Allen Common Coordinate Framework, a Coordinate Space is created, which redefines the zero point to be the top-left-front corner of the reference volume. Because some volumes were not defined using the brains of live animals, we further support Coordinate Transforms, which are affine transforms of an atlas and may optionally redefine the zero coordinate to a different position, such as Bregma, as shown here. (b) Sagittal and coronal slices are shown for the Allen CCF Coordinate Space and (c) for the Qiu et al. (2018) Coordinate Transform, demonstrating how the transform is tilted upwards, stretched along the AP axis and compressed along the DV and ML axes.

Integration with external hardware and software. Pinpoint exposes two application programming inter-faces (APIs) that allow for annotation and position data to be streamed to other applications and for manipulator positions to be echoed in the Pinpoint scene. We present three use cases for these external tools. (a) For researchers who want to improve their targeting accuracy during insertion, Pinpoint can send anatomical information about the final predicted target location of a probe to data acquisition software, such as the Open Ephys GUI or SpikeGLX. Users can compare these per-channel annotations against the electrophysiological signatures they observe on their probe to optimize their targeting. An example of the API channel annotation format is shown in the center box. (b) Pinpoint can also be linked to hardware micro-manipulators from Sensapex and New Scale, showing a live estimate of the current position of the probe during an experiment. This data can then be optionally forwarded to data acquisition software to display a real-time estimate of the probe’s position alongside the electrophysiology. (c) Finally, for users who want to maximize the efficiency of multi-probe recordings and minimize the potential for user error, Pinpoint offers a Copilot mode in which the insertion process is run by the software with minimal user intervention. (d,e) Examples of the Open Ephys and SpikeGLX graphical interfaces displaying probe annotation information.