Visual-Guided Solutions in Automated Zebrafish Larva Heart Micro-Injection

Micro-injection is one of the most efficient approaches to deliver foreign materials into zebrafish larva heart. Well-trained micro-injection technicians are required for such operation, and still the efficiency remains low and the work is laborious. As a result, automated micro-injection that is able to improve the efficiency and injection consistency is in great demand. Barriers like auto-focusing after changing of microscope magnification, accurate tracking of injection pipette, and zebrafish larva heart detection in three dimensions make the automation challenging. The proposed coarse-fine searching strategy and the depth of field optimized step lengths result in fast and accurate auto-focusing on the zebrafish and injection pipette. The injection pipette is tracked in real-time thanks to the proposed dynamical region of interest. Besides, the combination of y-Sobel filter and alpha-beta filter improves the noise resistance during pipette tracking, which helps to reduce the tracking error. With the proposed in-plane and out-of-plane heart detection algorithms, the zebrafish heart is detected in three-dimensions efficiently and the accuracy is promising. The success rate for automated zebrafish larva heart micro-injection reached 90% with the proposed visual algorithms.