Lumen Boundary Motion Analysis by Auto-correlation

We describe a method for extracting boundaries, such as the lumen of a vessel, from digitized intravascular ultrasound (IVUS) frames from a S-VHS video, together with other useful parameters for each frame. Autocorrelation of the lumen and other parameters, a timeseries analysis, shows underlying regularity in the variation of lumen area through the heartbeat, and systematic variation of the circularity (or eccentricity) with heartbeat. The heart rate as extracted from video image sequences agrees with direct determination during animal experiments within 1%. The methodology should be useful in extensions to physiological and pathological