For my thesis project, I developed an algorithm to automate the micropipette setup procedure for Robotic ICSI. These fragile tools are as thin as 5 or 10 microns wide at their tips and are often indistinguishable to the human eye. They cannot be seen under a microscope unless it is in focus in the area of focus.
Manual set up of a micropipette can be a lengthy procedure and is strongly dependent on the skill and experience of the operator. During an event where the micropipette tip collides with its surrounding objects, the pipette tip often breaks, which can be catastrophic in Robotic ICSI.
This new algorithm was inspired by an intuitive search algorithm, utilizes image processing techniques to successfully bring an out-of-focus and out-of-view micropipette into the center of the field of view and into focus in approximately 30 seconds. Tests have shown that this system performs with high accuracy, high reproduce-ability, and low failure rates
Check out the IEEE article here:
Locating End-Effector Tips in Robotic Micromanipulation
Locating End-Effector Tips in Robotic Micromanipulation. J. Liu, Z. Gong, K. Tang, Z. Lu, C. Ru, J. Luo, S. Xie, and Y. Sun. IEEE Trans. Robotics, Vol. 30, pp. 125-30, 2014
Check out the first page of my 4th-year research thesis below:
some sample images of the functions used in the algorithm:
Step 1: detect the object:
Step 2: locate the tip and bring it into focus:
Some statistics on its responsiveness: