Transforming Acquisition Statistics Into Contrast, Without Decisions, Thresholds, Or Templates
Fred S. Greensite · 2005
In standard medical images (where gray scale value is linearly related to measured signal value within some window), contrast is based on the metric in signal space. However, it is possible to derive an equally natural set of images from the samz acquisition in which contrast is based on the sign’al value distribution, is independent of the signal space metric, and involves no decisions, thresholds, or filters. Nominally, these image the “information content of each voxel”, and the procedure is an alternative image processing approach when segmentation is inappropriate because of uncertainity due to noise. Such images may provide significant contrast enhancement in situations of very low contrast on standard images. The method has a number of desirable theoretical properties including the fact that standard metric imaging emerges as a special limiting case.