Cardiologists use CAMRA to perform tempo-spatial analysis of the heart function

Edinburgh 12 November 1999In collaboration with two hospitals in the United Kingdom and one clinic in Finland, the Edinburgh Parallel Computing Centre (EPCC) designed the Cardiac Magnetic Resonance Analysis (CAMRA) software, which is based on Magnetic Resonance Imaging (MRI) technology. The CAMRA programme, developed under EC-funded ENTICE co-ordination, enables the accurate and reproducible measurement of cardiac function, essential to combating heart disease, one of the leading causes of death in western society. A version of the new high-speed semi-automatic analysis software is now freely available on the Internet for hospitals in Europe.


The major aim of the CAMRA software was to develop a semi-automatic method for the rapid analysis of short cardiac axis MRI images. The clinical acceptance of the programme depended on the robustness between observers and acquisitions performed at various centres using potentially different protocols. The evaluation of the CAMRA software is unique in the fact that multi-centre data acquisition and manual/automatic analysis has been performed. The results from the European multi-centre, multi-observer clinical trial have shown that the CAMRA based semi-automatic segmentation of the left ventricle endocardium provided good agreement between the average of the manual analyses and reduced the inter and intra-observer variabilities in a cohort of twenty patients.

Martin Graves, Clinical Scientist at Addenbrooke's Hospital, Cambridge, who was a major player in the development and validation of CAMRA explained that CAMRA is the first low-cost solution to offer high quality images without damaging the patient. It actually produces four-dimensional images, namely three spatial dimensions plus time as a fourth, integrating the various time intervals involved, charted to show a progression of movements in the heart. As a result, cardiologists are offered a good idea of how a particular patient's heart is functioning.

MRI forms an important development in the diagnosis and management of patients with cardiovascular disease. There are approximately 8000 MRI scanners in use all over the world, of which around 300 are in the European Union. Although these numbers are increasing every year, very few centres perform MRI analysis due to the lack of good, affordable analysis software capable of processing four-dimensional data. Usability of such software is restricted even further as programmes tend to use proprietary image formats and can only run on expensive UNIX workstations.

CAMRA targeted all of these problems, aiming to improve the productivity of cardiovascular MRI studies by providing fast, reliable data analysis through an intuitive and clinically useful user interface. As a multithreaded, scalable tool working on a low-cost single or multiprocessor PC-based platform which runs under Windows NT, CAMRA enhances the full potential of MRI, a non-invasive technology for medical imaging. It uses the DICOM (Digital Imaging and Communications in Medicine) standard to import images over Ethernet via TCP/IP, allowing its use with a range of scanners from different vendors. This ensures that CAMRA is a cost effective and vendor independent tool.

The software supports both manual and automatic segmentation, the latter employing an active-contour model or ACM based technique, together with facilities to enable rapid review and modification of contours. An additional feature allows the epicardial border and valve plane on short cardiac axis images to be quickly determined from co-registered long cardiac axis images. The result is a fast, reliable, reproducible and cost-effective data analysis software capable of quantitative cardiac MRI analysis.

Among the hospitals which have requested to utilise the CAMRA software are the Edinburgh Western General Hospital and the Hôpital Cardiologique Louis Pradel in Lyon, France. The CAMRA technical report is available at the Web site of the Edinburgh Parallel Computing Centre. For more information, please also read the VMW article Parallel computing enhances health care analysis.

Leslie Versweyveld

[Medical IT News][Calendar][Virtual Medical Worlds Community][News on Advanced IT]