Search
Optical Target Recognition
|
Biochip production is dramatically enhanced by a video system that automatically detects small patterns such as microelectrodes, nanowells, microcantilevers, etc. using a proprietary software extension. This enables the Nano-Plotter to precisely find positions and spot onto them (manual selection of points by clicking in the camera window is also possible). For very small targets, the image recognition procedure is enhanced by a fine control of the actually dispensed spots (yellow paper test, see below). For regular patterns like grids of pads the image recognition software works "out of the box" with any further adaptation. For arbitrary patterns like microlectronic chip surfaces two specific features will be used as aligner marks. This procedure a "teach in" procedure to get the software ready for unattended run. The example shows the spotting onto an electrochemical biosensor chip developed by the Fraunhofer Institute for Silicon Technology, ISIT, Itzehoe (with kind permission by ISIT).
|
||||
|
START -> |
||||
 |
 |
 |
||
| Larger Image... |
Larger Image... |
|||
|
1. The image recognition module of the Nano-Plotter software contains a set of algorithms to detect positions on various target types. On the ISIT chip, two rectangular alignment marks are found. These marks define a coordinate system and a spot layout on the chip can be set relative to these marks (the picture shows our ChipDesigner software to enter their coordinates). No reprogramming is required if the chip layout changes. |
2. The microscope camera is positioned above the chip and a video image is grabbed. |
3. The image analysis algorithm finds the alignment marks and and calculates all preset spot positions, transformed into XY-printhead coordinates. |
||
 |
 |
 |
||
| 4. The next sample is taken from a microtiter plate. The correlation between sample well and index number of the spot position can be freely assigned by writing a simple “transfer list†file. |
5. To ensure that the tip is working reliably, a stroboscope test is performed. A light emitting diode flashes at the piezo frequency, resulting in a still image. The pattern recognition software grabs an image and checks whether the droplet(s) meet the preset conditions (number and size of droplets, deflection angle). If not, the pipette is emptied and sample aspiration is repeated. |
6. To find out the exact dispensing position of the micropipette, the pipette spots onto a water-sensitive paper, producing a blue spot on a yellow background. |
||
 |
 |
 |
||
| Larger Image... |
Larger Image... |
Larger Image... |
||
| 7. An algorithm for low-contrast images detects the spot on the paper and calculates its XY-deviation from the theoretical position. |
8. Now the Nano-Plotter is ready to spot exactly onto the next chip position(s). |
9. When the microscope returns to the chip, the spotting result can be seen in a live video image and/or stored for documentation. Here, every second position was spotted (seen by the reflexions). Finally, the pipette is emptied and washed. After this, the next sample is processed (step 4) or the program terminates. |
||
| -> FINISH | ||||
