Siemens set to launch iterative recon CT software at RSNA 2009
Dose reduction will be among Siemens’ key messages at the RSNA meeting this year. Driving home this message will be IRIS (iterative reconstruction in image space), a proprietary algorithm that processes raw data acquired by CTs, according to André Hartung, Siemens vice president, CT Marketing and Sales.
Dose reduction will be among Siemens' key messages at the RSNA meeting this year. Driving home this message will be IRIS (iterative reconstruction in image space), a proprietary algorithm that processes raw data acquired by CTs, according to André Hartung, Siemens vice president, CT Marketing and Sales.
Early results from the use of IRIS prototypes indicate the new product could cut dose by up to 50% without compromising image quality. These experiences have generated data scheduled for peer-reviewed journal publication, possibly before the RSNA meeting.
Beta test sites have focused initially on evaluating the dose reduction capabilities of the iterative reconstruction software on Somatom Definition scanners that offer only a single source of x-rays. This group is composed of the Definition AS (Adaptive Scanner) family with 20-, 40-, 64-, and 128-slice versions. The Definition brand also covers dual-source configurations. Hartung expects IRIS to be available for all Definition products, including Siemens' dual-energy Definition Flash CT.
The software may be shipping on new CT systems from Siemens as early as spring 2010, Hartung said. An upgrade is being developed for installed Definition systems.
When applied to the Definition Flash, IRIS will allow users to cut dose beyond an already low 1 mSv for scans of the thorax, including the heart, according to Hartung. Meanwhile, results have been presented that indicate dual-source imaging can cut dose in selected cases, even without IRIS. These involve eliminating the need for precontrast images when contrast-enhanced studies are indicated.
Siemens will also feature at RSNA 2009 a CT Neuro Engine to assess stroke. The Neuro Engine processes data acquired using a volume perfusion technique. Such advanced applications can run on scanners with fewer slices, although they are best suited to superpremium devices that generate 256 or more slices per rotation.
Can CT-Based Deep Learning Bolster Prognostic Assessments of Ground-Glass Nodules?
June 19th 2025Emerging research shows that a multiple time-series deep learning model assessment of CT images provides 20 percent higher sensitivity than a delta radiomic model and 56 percent higher sensitivity than a clinical model for prognostic evaluation of ground-glass nodules.
