Novel dual energy multi-detector computed tomography (DECT) image processing software can accurately characterize urinary calculi composition.
Development of the software (Syngo Multi Modality Workplace, Version VE25A, Siemens Medical Solutions, Forchheim, Germany) for determining stone composition represents a collaborative research effort involving members of the division of urologic surgery and department of radiology at Duke University Medical Center, Durham, NC, including senior authors Glenn Preminger, MD, and Daniel Boll, MD.
The software uses a pixel-by-pixel analysis to generate dataset gray-scale-encoding ratios of relative differences in attenuation values of low-energy (80 kVp) and high-energy (140 kVp) peaks for each stone read. Results from an in vitro pilot study showed that in graphic analysis, where the renal calculi were plotted based on the calculated attenuation identifiers, stones of different chemical composition sorted into distinct clusters.
A total of 82 stones were read. The spectral analysis showed the represented stone types included uric acid, struvite, calcium oxalate monohydrate, calcium oxalate dihydrate, calcium phosphate, brushite, and ammonium urate. From the radiology standpoint, 51 of the 82 stones were considered pure based on a definition of being at least 70% single crystalline composition.
Software helps distinguish stone types
In graphical display, there was significant overlap between the various stone groups as assessed by the radiologist's readings. However, with the implementation of the post-processing algorithm, stone groups were successfully separated into sharply demarcated clusters.
"Former attempts to determine stone composition by analysis of CT images have only been partially successful," said Dr. Zilberman.
"Implementation of this improved algorithm would allow noninvasive, pre-procedural determination of stone composition that could help guide management decisions about whether shockwave lithotripsy or endoscopy is appropriate, and in certain situations, enable initiation of medical treatment without performing a metabolic evaluation or surgery first."
An article reporting results of the study is currently scheduled for publication in the December 2010 issue of the Journal of Urology.