Title: Strongly Interactive Medical Imaging Algorithm Design


			     James Williams
		Head, Imaging & Visualization Department
		       Siemens Corporate Research
			 755 College Road East
		          Princeton, NJ 08540
	         Email: james.williams@scr.siemens.com


Abstract:
When a skilled user is presumed to be in the loop for a given
diagnostic task, the nature of the algorithms is significantly different
from those used in offline or automated tasks.
Initialization, guideance and correction of the techniques must be
intuitive. This leads to a class of algorithms which must be reactive,
but can expose parameters in areas where uncertainty can
be estimated and visually depicted. 
Examples will be given in the cardiovascular & general body imaging
domains with applications in segmentation and quantification.