Medicsight announces market release of ColonCAR software with significantly improved reader performance in published CT colonography study

London 12 December 2006Medicsight plc, a subsidiary of Medicsight Inc. has published a study in the December 2006 issue of Gastroenterology that showed Medicsight CAD (computer-aided detection) significantly improved reader interpretation of CT colonography. The paper was published with a directed editorial. Medicsight PLC's ColonCAR 1.2.1 has been validated through Memo to File. Medicsight's ColonCAR product is ready for market, based upon information available under a current Memo to File designation.


Medicsight PLC recently announced the market release of Medicsight ColonCAR 1.2.1 in the USA and European Union. ColonCAR 1.2.1 is an image analysis software tool designed to be used with CT colonography (virtual colonoscopy) to assist radiologists in searching for and measuring potential colorectal polyps. Medicsight introduced the product at the 92nd Scientific Assembly and Annual Meeting of the Radiological Society of North America (RSNA) conference in Chicago.

Medicsight's ColonCAR (Computer Assisted Reader) uses a series of filters deployed against image data from CT colonography studies. The filters highlight spherical areas as small as 5mm or the size of a small pea, allowing radiologists to examine these areas in 3D in order to judge their potential as true polyps.

Radiologists are also able to manually highlight any irregularities for closer inspection. Once suspected polyps are found, the software can determine the shape and features, precisely identifying the boundaries and showing the lesion in 3D with automatic diameter and volume measurements. This allows the radiologist to accurately review and monitor polyps over time.

"We are really excited about bringing this innovative and leading product to market and the benefits it will mean for radiologists", stated Etienne Vanderstokker, CEO of Medicsight PLC. "We are confident that the integrated solution will create ideal work flow and optimize reading times for end user productivity, thereby giving Medicsight the opportunity to capitalize on the ever increasing customer demand for CAD."

Colorectal cancer is the third most common cancer found in men and women in the United States, according to the National Cancer Institute. American Cancer Society statistics on colon and rectal cancer estimated 148.610 new colorectal cases to be diagnosed in 2006, accounting for almost 14 percent of annual cancer deaths.

The Medicsight ColonCAR software will be integrated within leading imaging 3D advanced visualization products from Vital Images Inc., TeraRecon Inc., Viatronix Inc. and 3mensio Medical Imaging BV. Medicsight's ColonCAR software will also be available for order through a new medical portal,, which was launched at the RSNA show in Chicago.

The study by Halligan S., Altman D.G., Mallett S., Taylor S.A., Burling D., Roddie M., Honeyfield L., McQuillan J., Amin H., and Dehmeshki J., titled "CT colonography: Assessment of radiologist performance with and without computer-aided detection" and issued in Gastroenterology 2006 (in press); ePub November 6th, investigated the effect of CAD assistance on the performance of 10 readers who were "non-expert" in the interpretation of CT colonography. While there have been several studies of CAD performance in isolation, this was the largest study to investigate the effect of CAD on readers' interpretation of CT colonography. The study was performed by Professor Steve Halligan and colleagues from the University College Hospital London and the Centre for Statistics in Medicine, Oxford, United Kingdom.

Each participant read 107 CT colonography cases: 60 patients had 142 polyps and the remaining 47 had normal colons. The cases were first read without the assistance of CAD and then two months later the studies were read with the assistance of CAD (Medicsight ColonCAR). The study found that readers detected significantly more polyps with CAD than without. For example, with CAD 41 (68 percent) of the 60 patients with polyps were more likely to be detected by readers.

This effect was mostly a result of improved detection of polyps measuring 9mm or smaller in maximal diameter. Importantly, specificity was not adversely affected by CAD - while sensitivity increased significantly in 7 readers, specificity declined significantly in only one.

The study also investigated a concurrent read paradigm, where the CAD annotated and un-annotated images are presented to the reader simultaneously. The investigators found that this approach shortened reader interpretation time significantly when compared to CAD unassisted interpretation: by 1,9 minutes for patients with polyps and by 2,9 for those without. Overall, 9 or 90 percent of the readers benefited when using CAD, either due to increased sensitivity, decreased interpretation time, or a combination of both.

Medicsight is a pioneer in the development of enterprise-wide computer-aided detection (CAD) and Computer Assisted Reader (CAR) software used by the imaging community to aid in earlier detection of disease. Medicsight's unique software solutions have been validated using some of the world's largest and most population-diverse verified CT scan databases to help clinicians identify, measure, and analyse suspicious pathology of the lung and colon.

The company's products include ColonCAD and ColonCAR, the first technology available for CT colonography, and LungCAD and LungCAR. Medicsight continues to develop CAD software for a variety of conditions that can help in the early detection of disease and ultimately improve patient outcomes. Headquartered in London, Medicsight plc employs more than 35 people and also has offices in the United States and Japan. More company news can be found in the VMW October 2006 article Medicsight to support continuing medical education in virtual colonoscopy.

Leslie Versweyveld

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