Search Contact information
University of Cambridge Home Department of Engineering
University of Cambridge > Engineering Department > Machine Intelligence Lab > Medical Imaging Group

Graham Treece

[ Introduction | Teaching | Research | Publications | Software | Videos | Personal ]

Click here for other medical imaging projects offered by Graham Treece.

F-GMT11-2: Imaging of coronary plaque using ultrasound elastography.

Build up of plaque in arteries is a problem since it can restrict blood flow, but more importantly the plaque can rupture, potentially causing blockage elsewhere. (Image from Intravascular Ultrasound (IVUS) is a technique for imaging arteries from the inside using a catheter. But such images don't necessarily tell you whether a plaque is vulnerable to rupture or not.

Rupture of coronary plaque can cause blockage of arteries, resulting in a myocardial infarction (often called a heart attack). But just because plaque exists does not mean it will rupture: much plaque is stable. Hence the detection of vulnerable plaque is an important clinical need. It is not easy to determine whether a plaque is likely to rupture, but there is evidence to link this to the mechanics of the plaque, i.e. how stiff it is and what are the resulting stress and strain distributions in the plaque.

Ultrasound elastography is a technique which can image strain distributions in vivo by using ultrasound. We have recently developed techniques for producing high quality, clear, and good resolution images using this with linear array ultrasound probes (that you can hold in your hand and use to image a patient from the skin surface). It is also possible to image within an artery, by using specially designed ultrasound catheters, and elastography has already been investigated in this context to look at vulnerable plaques.

The aim of this project is to see whether some of the recent developments for elastography using linear array ultrasound probes can be applied to intra-vascular imaging for plaque detection. This is quite a hard problem: apart from anything else, the geometry is completely different in this case. The aim is to produce an imaging tool which is robust enough to be used in a clinical scenario.

This is an algorithmic development / mechanics / software project, so experience of writing software is essential, though some of the development could also be done using Matlab. The project involves some collaboration with the cardiovascular medicine research group at Addenbrooke's Hospital.

This is an example of what can be done using normal linear ultrasound arrays: here imaging a stiff spherical object in a soft background. This is the raw strain data which forms the basis of an elastogram. Normalisation is the process of guessing what the stress distribution was, which results in a much more even background and a better estimate of how noisy the measurements were. The measurement noise can now be used to mask out (shown in red here) some of the image. However, it can also be used to optimally combine sequences of images, to give really stable, high quality elastograms.
© 2005 Cambridge University Engineering Dept and Graham Treece .
Information provided by gmt11