Managing the Flow in Radiology

Tool Promises to Ease Dementia Diagnosis
Software: Two that Are New
New IPIX Staves Off Germs, Dust

Tool Promises to Ease Dementia Diagnosis

Interpreting FDG-PET brain-scan images requires extensive time and skill, especially when locating signs of dementia in the early stages of neurodegenerative disease. But with a new decision support software tool for automatic evaluation of dementia using PET image data, the process could become s imp l e r a n d quicker, improving patient care.

The decision support software tool, developed by Philips, has shown positive results in retrospective studies. The software is designed to analyze PET brain-scan images, combine them with MRI scans, and distinguish between signs and symptoms of brain diseases. While early differential diagnosis of dementia-causing conditions, including Alzheimer’s, frontotemporal dementia, and Lewy body dementia, is essential to maximize drug efficacy and prevent loss in quality of life, the process is painfully subjective and difficult. The decision support software tool claims to automate and speed this process, detecting the characteristic patterns of each of these differential diagnoses, then quantifying the degree to which they resemble reference images of known dementia sufferers.

The University Medical Center Hamburg-Eppendorf in Hamburg, Germany, has conducted usability and feasibility studies of the software over the last year. It used libraries of FDG-PET brain-scan images already independently interpreted by experts to test the tool. The decision support software—programmed to differentiate between brain scans clear of dementia symptoms, scans indicating Alzheimer’s disease, and scans characteristic of frontotemporal dementia—corresponded with the existing interpretation of 83 patients more than 98% of the time.

“The results of the study have been truly excellent,” said Ralph Buchert, MD, of the Department of Nuclear Medicine at the University Medical Center Hamburg- Eppendorf, in a statement. “With most steps performed automatically, operating the software is very straightforward and only adds a couple of minutes to the procedure time. When we made it available to referring physicians who were not skilled in reading FDG-PET and MRI images, they were able to analyze images and interpret the results within a few minutes.”

In a similar study at Austin Hospital in Melbourne, Australia, the software achieved better than 80% accuracy in a pool of 48 patients. The researchers tested the software in the more difficult area of differentiating between the scans of patients without diseases, patients with Alzheimer’s, those with frontotemporal dementia, and those with Lewy body dementia.

The software works through a series of three steps. In the first step, the patient’s brain-scan image is spatially normalized, selecting, rotating, and scaling the appropriate image slice to align it with a standard template. Next, the normalized image is compared with a library of brain scans without disease, identifying hypometabolic regions in the patient’s brain. Finally, these hypometabolic regions are compared in their size, shape, and distribution against a set of disease-specific patterns for each type of dementia. The software also quantifies the degree to which the patient’s scan matches each disease-specific area, in the form of a percentage.

The software is designed to improve the often difficult existing process of diagnosing dementia. While FDG-PET brain scans track uptake of glucose in the brain, pinpointing the areas of reduced glucose uptake that indicate potential dementia-causing conditions, the final scans can be a series of subtly different shades of gray. Radiologists can have difficulty determining if disease exists, and especially distinguishing between Alzheimer’s and Lewy body dementia, which originate in the same area in the back of the brain.

To further test its software and develop the tool as a reliable, automated method of differential dementia diagnosis, Philips Research will continue its studies, conducting additional research involving the software in collaboration with the University of Washington in Seattle.

—Amy Lillard

Software: Two that Are New

INFINITT North America, a privately held company specializing in the production of Web-based RIS and PACS solutions, has two new technologies to show off featuring its Xelis line of products.

Xelis software modules provide accelerated data loading and clinical tools for enhanced 3D viewing of large-volume data sets.

The Xelis products are software modules designed to provide accelerated data loading and clinical tools for enhanced 3D viewing of largevolume data sets while allowing Webbased access from any Internet-connected PC. The Xelis line of products can also be fully integrated with INFINITT’s own Webbased PACS.

INFINITT debuted its Xelis-Lung Solution in late October. The Xelis-Lung Solution uses postprocessing techniques such as volume rendering to help improve nodule detection from multirow detected computed tomography (MDCT) data. It includes automated nodule detection for solid type and ground glass opacity (GGO) type nodules along with automatic 3D volume and statistical measurements, helpful in nodule classification.

The system also allows three viewing modes, according to INFINITT: one for manual-detected nodule review, followed by manual-detected nodule review followed by auto-detected review, and one for concurrent manual and automatic detection review. To enhance workflow, the software also includes a user-friendly segmentation tool, an annotating or labeling tool, and automatic job save to store working context.

INFINITT’s second technology featuring Xelis is the Xelis Cardiac, which debuted at the 2008 Radiological Society of North America meeting. Xelis Cardiac uses postprocessing techniques including volume rendering to improve visualization and analysis of coronary vessels from multislice CT data. Xelis Cardiac includes the Vessel Analysis software package to help improve radiologists’ productivity and accuracy. The package features singleclick automatic vessel tracing, MPR view, Angio-emulation view, various segmentation functions, automatic rib cage removal, Compare Curved Planar Reformation (C-CPR), stenosis and soft plaque analysis, as well as automatic calcium scoring.

INFINITT North America is headquartered in Phillipsburg, NJ, and has been involved in the development of image and information management technologies since it was founded in 1993. The company’s goal is the production of affordable, comprehensive Web-based solutions that can operate on a single platform, simplifying workflow. Currently, INFINITT’s software is used in more than 1,500 health care institutions in 14 countries.

—Ed Wilson

New IPIX Staves Off Germs, Dust

“The compact size of the IPIX and the overall design of the all-in-one system also make the IPIX an excellent choice for OEM applications where the IPIX may be used for control or monitoring of diagnostic or therapeutic equipment,” Davis said. “All connectors are located in a central connector block, making it easy to incorporate the IPIX into an existing medical device.”

Additionally, the IPIX is designed to accept a variety of input power, which enables the Intel Core 2 Duo-based IPIX to replace outdated control or monitoring systems. In order to incorporate the IPIX into new or existing equipment, users can employ industry standard VESA mounting solutions.

“The IPIX makes it easy to bridge the gap between the medical worker and the IT department, providing a system with medical needs in mind while providing an advanced system that meets the demands of the IT staff as well,” Davis said, adding that optional touch screens encourage user-friendliness and advanced processor and wireless capabilities make for easy integration into the IT setting.

The IPIX is available with a 17-inch SXGA color display or a 20-inch UXGA display. The latest version is the fourth-generation all-in-one computer from Datalux. It comes with a standard 3-year, return-to-factory warranty.

—E. Sanchez