Automated placement of stereotactic injections using a laser scan of the skull

TL;DR

This paper presents a software tool that uses 3D laser scans of mouse skulls to improve the accuracy of stereotactic brain injections, reducing errors compared to traditional landmark-based methods.

Contribution

The study introduces a novel method employing 3D skull scans and point cloud registration to enhance precision in stereotactic injections in mice.

Findings

Registration error less than 100 microns

Injection targeting error approximately 500 microns

Potential for widespread automation in stereotactic procedures

Abstract

Stereotactic targeting is a commonly used technique for performing injections in the brains of mice and other animals. The most common method for targeting stereoscopic injections uses the skull indentations bregma and lambda as reference points and is limited in its precision by factors such as skull curvature and individual variation, as well as an incomplete correspondence between skull landmarks and brain locations. In this software tool, a 3D laser scan of the mouse skull is taken in vitro and registered onto a reference skull using a point cloud matching algorithm, and the parameters of the transformation are used to position a glass pipette to place tracer injections. The software was capable of registering sample skulls with less than 100 micron error, and was able to target an injection in a mouse with error of roughly 500 microns. These results indicate that using skull scan registration has the potential to be widely applicable in automating stereotactic targeting of tracer injections.