Automatic Microinjection Assistance System Based on Deep-Learning Visual Positioning

With the development of life sciences, the requirements for biological cell injection technology are getting higher and higher. Nowadays, most of the injection work is done manually, which is not only time-consuming and laborious, but also difficult to maintain a stable injection efficiency. Some automatic microinjection systems have been developed to improve efficiency, but it is essential to accurately identify and position cells and injection needles. This paper proposes a method for visual positioning work using a hardware device in combination with a deep learning target detection algorithm. The microinjection task is then completed by the integration of visual positioning and motion control via the upper computer control module. This microinjection assistance system is guided by visual positioning to reduce the difficulty of automatic microinjection and increase injection efficiency. Finally, the effectiveness of the microinjection assistance system is verified by using zebrafish embryos as an example, which lays the foundation for the subsequent development of automatic and intelligent cell microoperation systems.