Algorithm for the co-registration of virtual histology intravascular ultrasound (VH-IVUS) images for tracking the progression or regression of plaque during coronary artery disease.
- Allows for regional assessment of plaque development and progression unlike traditional techniques that analyze plaque progression as one quantity for each paired VH-IVUS.
- Co-registration occurs in almost real time.
- Provides a method for quick, easy, and accurate comparisons of IVUS images.
IVUS is considered the gold standard for intravascular imaging. The global market for intravascular ultrasound is forecast to reach $698 million by the year 2017. The IVUS procedure can be used with a single use imaging catheter or in a reusable imaging system. Although this IVUS procedure is widely used, due to catheter rotation during wire pullback and image acquisition, longitudinal images (acquired over time) cannot be rotated so that circumferential bases do not coincide in each image, i.e. â€œ3 oâ€™clockâ€ in one image does not correspond to â€œ3 oâ€™clockâ€ in the corresponding image. These traditional techniques can only analyze plaque progression as one quantity for each paired IVUS image even though plaque development and progression occurs heterogeneously around the artery circumference.
Because co-registration is not available in these IVUS machines or in any of the software for these machines, researchers from Emory University and Georgia Tech have developed an algorithm that allows circumferential co-registration of longitudinal IVUS images. Image processing techniques including gradient filtering, thresholding, and size exclusion are applied to each image to extract the inner and outer boundaries of the internal and external elastic lamellae, respectively. Thickness values are calculated around the vessel circumference and a normalized cross-correlation is determined. The baseline image can then be rotated to correspond with the follow-up image. Unlike traditional technologies that assess plaque development and progression as one quantity, this technique allows for assessment on a point-by-point basis allowing for analysis of the regional progression of coronary artery disease. This technology could either be bundled together with an imaging system or sold as an add-on.
Algorithm has been tested with co-registered image sets from patients.
Publication: Timmins et al., 2012. ASME 2012 Summer Bioengineering Conference, Jun 20-23, 2012.
View our Breakfast Club video
Review our Breakfast Club PowerPoint