Automatic detection of cell contact for microinjectionIntroduction Microinjection tests are done in many areas such as sys-temsbiology (functional genomics), pharmacology and-toxicology. Different kind of substances like geneticma-terial or fluorescence dyes are injected into the cellsin order to manipulate and monitor the biochemistry ofthese specific cells. However, microinjection of largeamounts of cells is a tedious and time consuming task VTT has developed a machine vision system for the contact detection between the microinjection needle and the cells in order to ease the operating of the mi-cromanipulation and make the automatic microinjec-tion possible. A computer-controlled micromanipulator developed at Tampere University of Technology was used for testing the injection capability.  Graphical tools development for cell inspection software A simple graphical tool was created for marking and saving the cell locations from a live image and the initial position of the needle tip for mapping these positions to the controller of the manipulator. The cell locations can be marked or deleted from a live image and they are graphically visualised in the user interface. The operator can manually select the cells one by one for detecting possible contact or send all the cell locations to the micromanipulator for automatic detection. Visual look of the interface can be seen in Figure 1. Contact detection between the microinjector needle and the cell membrane During injections, the phase contrast lighting was utilised. When the operator of the micromanipulation system is doing the injections, the best possible situation for detecting the contact is that the cell is bent or sufficiently moved by the needle resulting in a change of luminosity at the contact point. Therefore, the contact can be seen as a bright point at the tip of the needle as the tip bends the membrane (see Figure 2). Additionally, events like the movement of the nuclei and the movement of the cell itself when the needle touches the cell, also gives clues of the contact. However, in most cases, the changes that happen during a contact are very small and may be overrun by the motion and focusing of the needle so that an inexperienced operator is unable to observe the contact. If the contact is missed, the needle runs against the bottom of the dish and the tip of the needle will break down. By using machine vision techniques with the proper image processing, the contact point can be emphasized to the operator and the contact can be automatically detected for controlling the micromanipulation.  Summary VTT has developed machine vision based methods for detecting the instance when a computer controlled microinjection needle touches the membrane of a selected living cell in an incubation vessel. Such methods can be used for automatic selective microinjection of cells or aspiration of a single cell contents for metabolic analysis. The work has been carried out in a consortium with partners from Tampere University of Technology, Automation and Control Institute, and Tampere University, Cell Research Center, and companies ChipMan Technologies, FIT Biotech, Karyon, Medicel, Orion Pharma, and SPR Veripalvelu.
|
