Study On Key Technologies Of Automated Microinjection System Based On Micro-robot

The traditional cell microinjection relies on expensive operating devices and long-termed trained skilled clinicians, and requires the operators have a good physiological condition and stable psychological qualities. However, the subjective psychological fluctuations and physical fatigue still can not be avoided, leading to poor repeatability and low success rate. To overcome the defects caused by manual operation, we design and researcha a small size, low cost and easy operated microinjection operation robot(referred to as micro robots hereinafter), instead of the traditional manual cell microinjection. So it has great significance to achieve cell microinjection automatically.Aimed at the issue of micro robotic automation cell injection, firstly, we need to design a micro robot with mutiple degrees of freedom which is suitable for cell injection. Secondly, the kinematic model of injection pipette must be built and cell attitude adjustment must be analyzed before cell injection. Lastly, each stage of cell injection is segmentated by using image segmentation algorithm. And the segmentation result is used to lead the micro robot to achieve cell injection automatically. Therefore, in this paper we studied four aspects: micro robot design and reseach, analysis of non-touched cell adjustment and injection pipette motion, segmentation for the cytoplasm edge of human oocyte during microscopic injection process, segmentation for cytoplasm, first polarbody and injection pipette in human oocyte during microscopic injection process. The main contents and innovation are as follows.1. Design and research of the 3 degree of freedom hybrid driven micro robotFor the diffficulty and high cost of traditional manual operation of micro injection, we design a cell injection oriented micro robot which is driven by piezoelectric ceramic and voice coil motor. Firstly, the cell microinjection system is based on micro robot. Secondly, a smart mechanical structure of micro robot body(planar motion module and longitudinal movement module) is presented. And then the design principle of the signal generator and power amplifier of micro robot driven system is introduced highlightly. Then, we introduce the principle of planar inchworm motion and vertical micro-motion of micro robot, and elaborate the straight line walking, left and right turning, in-situ rotation and micro displacement in vertical direction. Finally, a user interface is shown and automatic microscope focusing is presented.2. Analysis of injection pipette motion and cell attitude adjustmentWe have to prepare for the motion of ISCI injection pipette and cell attitude adjustment before cell injection. Firstly, establish kinetic model of injection pipette whose parameters are identified by least square method and design LQR(linear quadratic regulator) to control the position of injection pipette under microscopic view. Secondly, stress analysis model of cell is established, micro flow field and rotary motion of cell is simulated by using FLUENT dynamic mesh parameters. Then, study 3 impact factors on the micro flow field that including distance between two micro pipettes, inner diameter of micro pipette and the velocity of fluid from micro pipette. Finally, design a solution of cell attitude adjustment.3. Cytoplasm segmentation research base on the edge constraint level set methodAiming at cytoplasm edge segmentation of human oocytes in the microinjection process, a level set model of cytoplasm edge segmentation base on the edge constraint method is proposed. Firstly, image preprocessing is implemented. According to the characteristics of shape difference, the outline of cell and injection pipette is extracted by the image enhancement filtering and morphological method. In addition, it is applied in the initial position of level set evolution. Secondly, segmentation model of variational level set base on edge constraint is established. The edge constraint condition is established by utilizing local gray distribution characteristic in cytoplasm edge and combined with variational level set theory to improve energy functional model. Finally, each stage of deformative cytoplasm edge is segmented in the cell injection process. Namely, segmentation research is carried on contacting oocytes, critical piercing of oocytes、microinjection and injection pipette retreat, respectively.4. Segmentation of oocyte puncturing process based on fuzzy clustering and level setAims at solving the segmentation problem that seperating injection pipette, cytoplasm and first polarbody of human oocyte during oocyte puncturing process, a segmentation model combined fuzzy clustering with level set is proposed for the oocyte puncturing process. Firstly, extract the target of interest using skeletal morphology according to different target shapes. Secondly, the preprocessing image which has large noice or fuzzy boundaries, is clustered coarsely by the feature of local grey value and local variance. Thirdly, the initial location of level set and the evolution parameters are set up based on fuzzy clustering and realize the segmentation of cytoplasm,first polarbody and injection pipette inside cytoplasm. At last, a complete image of the needle is marked by direct line fitting.