Out-licensing

Optimal Whole-Heart Ultrasound Imaging

Introduction/Background

Recently developed technology has allowed, for the first time, the acquisition of 3 dimensional ultrasound echo images of the heart in real time. This new imaging modality opens a wide range of possibilities in echocardiography. However, it is not currently possible to scan the whole adult heart in a single acquisition, hence the development of tools to combine acquired images is of great importance. Simple techniques currently used to combine the images, which work by taking the mean or maximum intensity at each pixel (or voxel), can result in the reduction of information content. The difficulty in combining the ultrasound images, while still retaining the original information, limits the visualization and quality of image analysis that can be performed.

Aims/Hypothesis

A solution would be of great benefit to medical ultrasound imaging.

Research

A new method of combining a number of images of a common object has been developed at the University of Oxford. The key features of this technology are the use of a feature measure protocol that identifies key features of interest in each of the original images in order to retain important information in the combined image, the use of knowledge of the ultrasound acquisition process to identify the images that better characterize the scanned structures. This invention provides a way to fuse ultrasound images (2D, 3D, 2D+T, 3D+T) taken from different acoustic windows or views of the object to define a new image and relative weights of the images are estimated that result in a combined image in which the information content from the individual images is maximised.

These features combine to give a method that delivers a resultant image that is of better quality for subsequent visualisation and analysis. The resulting combined image may be used in a variety of manners including general tasks such as display, segmentation or tracking. However, the improved quality of the images facilitates their application to more complicated tasks such as object recognition or alignment, for example in image-guided interventions, surgery and therapy.

Conclusion

We have developed a new method for combining ultrasound images to produce a single, optimal data set, whilst retaining the quality and content of the original images.

Relevance/Opportunity

Please contact the Isis Project Manager to discuss this further quoting reference no. 2586.

Registered

This work is the subject of a patent application