| | Re: cDNA Microarray storage, storage issues & compression solutions
In general, all microarray images are grayscale images (with the color coming from whatever filter is placed in front of the sensor at the time the image was taken). Multiple grayscale images representing different colors are taken using different filters. As for the filetype, that will depend on the system. Scanners that work with 'homebrew' printed slides will probably generate TIFFs (probably 16-bit), but the much more popular full commercial systems (like that of Illumina or Affymetrix) will store the image files in a proprietary compressed (lossless) format. It is generally stored as two TIFF images. Each image is stored as an uncompressed 16-b gray-level bitmap, with each pixel taking a value between for example; 0 and 65 535.
Data integrity and consistency; If a data break down occurs half way through the process, a table could have been deleted but remains registered in the catalogue of the database system - very difficult to update this.
Database standards; (Depending on the size of the array and the resolution of the scanner, images may require from a few megabytes to several tens of megabytes of storage)
There a five published methods for the lossy and / or lossless compression. Although different, they share nevertheless, a common structure: they perform griding followed by segmentation and then encode the background and the spots separately. The segmentation information is also included.
For example, a compression method based on arithmetic coding driven by a 3D context model. The image is compressed on a bit-plane basis going from the most significant bit-plane to the least significant bit-plane. Encoding stopped if an average of more than 1 bit/pixel is obtained. The remainder bit-planes are sent uncompressed.
Hope this helps!