|
Area (or "2D") scanning systems are comprised of a microscope, a conventional CCD digital camera, a motorized X-Y stage, and a Z-stepper. It can look like a standard microscope or can be integrated into a “closed box” design. When the slide is loaded a digital image of the field-of-view is captured (often referred to as a "tile"), then the slide is moved to the adjacent field-of-view to repeat the digitization process until the whole desired region of the slide has been scanned. The tiles produced by this method may then be "stitched" together by software which aligns each tile to create a large mosaic image.
Using a typical microscope camera with a resolution of 1280 x 1024 pixels, approximately 800 tiles are needed for a tissue section area of 15 mm x 15 mm if a resolution of 0.46 microns per pixel is to be achieved. If the desired resolution is 0.23 microns per pixel, then four times as many tiles have to be digitized.
Due to the "stop-and-go" scanning method, image acquisition is limited by mechanical considerations and inherently somewhat slow. Also, system calibration and the precision and lifetime of electromechanical components are critical.
Line scanning systems
These systems employ a linear array light sensor. The slide is scanned in a continuous manner by a linear movement perpendicular to the sensor axis. Using the example of above with a slidea area of 15 mm x 15 mm at a resolution of 0.23 microns per pixel, a linear array sensor with 4096 pixels would produce approximately 15 image "lanes". This greatly reduced number of adjacent scanned areas simplifies the process of tile alignment. Moreover, due to less severe mechanical constraints a faster scanning speed can potentially be achieved.
However, a fundamental performance trade-off remains: faster scanning speeds mean shorter light integration times on the sensor's pixels. This means that effectively the available scanning speed is limited by the lower sensitivity of an ordinary linear array sensor. Also, fluorescence operation may be unattainable.
The benefits of TDI
While the scanning operation of the NanoZoomer is linear, there is a big difference between an ordinary linear array sensor and a TDI sensor. A TDI sensor is not a linear sensor but actually a 2D sensor that is operated in a special way. You can think of it as a stack of many linear sensors, operating in synchrony.
In the NanoZoomer NDP scanner the TDI sensor acts in a similar way to 64 synchronized linear sensors. This keeps all the advantages of linear scanning, but yields a sensitivity 64 times higher than that of a single sensor, without compromising the scanning speed!
The benefit is a greatly enhanced image quality while maintaining very high scanning speeds. In addition, the sensitivity is high enough to scan fluorescence slides at similar speeds. | 
