< Previous page

NIH Image computes the minimum and maximum pixel values of 16-bit images and uses this information to linearly scale to 8-bits (1-254). Check Swap Bytes when importing 16-bit images from "little-endian" systems, such as an IBM-PC or VAX. If Calibrate is checked NIH Image automatically sets up a linear density calibration function to provide an approximation of the original 16-bit pixel values. The approximated values should not vary from the original 16-bit values by more than (Max-Min)/254, where Min and Max are the minimum and maximum 16-bit values. If Fixed Scale is checked then automatic scaling is disabled and the 16-bit data is linearly scaled from Min-Max to 1-254. You can redo the 16-bit to 8-bit scaling at any time using the Rescale command. Note that the Calibrate feature is disabled when importing a 16-bit stack and Fixed Scale is not checked.

Use the Import command to scale 16-bit images from medical scanners to the 8–bit format required by NIH Image. For example, to import raw MRI scans from a GE SIGNA scanner, set width and height to 256, select 16-bit Signed, and set Offset to 14336. The macro "Import GE Signa Files" (in the file Stacks) will automatically import a series of raw GE SIGNA files.

Check Open All to import all of the images in a folder. The images must all have the same width, height, etc. Open All does not work with imported text files and look-up tables.

Check Invert to invert imported 8-bit images.