Several members of the Kitware team attended RSNA from November 27-December 2. In addition to meeting with collaborators and exploring the exhibits, we were actively involved in demonstrating our newest work, teaching workshops and participating in exhibits.

 

 

 

On Monday, Kitware taught two courses, “Open Source Applications for Medical Imaging Research” and “Introduction to Open-Source Software Libraries for Medical Imaging”. These courses had sold-out audiences with over 100 attendees and received high acclaim.

 

 

In the open-source applications course, Rick Avila, Wes Turner, Julien Finet, and Patrick Reynolds presented.  Topics presented included VolView, Midas, KiwiViewer/VES, and Slicer.  For the VoView tutorial attendees viewed and analyzed medical images donated from the “Give-A-Scan” project for lung cancer research.  The tutorial covered basic image viewing functionality, and introduced the idea of using and developing VolView extensions for additional analysis capabilities. For the KiwiViewer tutorial, attendees saw how VTK is being extended (via VES) to support mobile and web-based applications. The power of Midas was demonstrated for collecting, processing, and sharing data in a manner that fits with medical imaging research workflows. The course also featured the recent 4.0 release of Slicer and illustrated its new Qt-based user interface, improved speed, and increased ease of use.

For the open-source libraries course, Patrick, Julien, and Rick presented several open-source software libraries used in medical imaging, including the Visualization Toolkit (VTK), the Insight Segmentation and Registration Toolkit (ITK), Midas, and VES. The course addressed managers, clinicians, researchers and scientists who direct and conduct the development of new medical image analysis and display applications. Highlighting the value of these open-source toolkits in medical imaging, we discussed the rigorous software processes used to ensure stability in the toolkits and specific applications of the toolkits to medical imaging. Participants were engaged with live demonstrations of coding of SimpleITK and new features of VES.

 

 

In addition to the courses we taught, we also had a presence in the Quantitative Imaging Reading Room (QIRR) with two exhibits. The first exhibit, “3D Slicer: An open-source software platform for segmentation, registration, quantitative analysis and 3D visualization of biomedical image data,” was a collaborative exhibit with the National Alliance for Medical Image Computing (NA-MIC).  At the exhibit, Julien Finet introduced translational clinical researchers to the capabilities of the 3D Slicer software at meet-the-experts sessions. A team of 3D Slicer experts ran hands-on demonstrations with sample datasets or, where appropriate, data provided by attendees. The demonstrations included volume rendered head, thoracic and abdominal CT scans, MRI-based topographic parcellation of a human brain, PET/CT quantitative assessment of tumor response, image-guided prostate interventions, longitudinal analysis of meningioma growth, white matter exploration for neurosurgical planning using Diffusion Tensor Imaging tractography, and registration and segmentation strategies for follow-up on cases of Traumatic Brain Injuries (see our RSNA page on the Wiki). This was a timely exhibit as it coincided with the official announcement of the Slicer 4.0 release, a major new release in the 3D Slicer series (see our earlier post on this topic.)

 

 

The second exhibit, “Understanding and Improving CT Image Quality With Automated Pocket Phantom Technology,” presented research on our 2nd generation CT pocket phantom. CT volumetric performance is highly influenced by fundamental acquisition properties, such as resolution. By combining this small form factor, accurate artifact with automated detection and measurement algorithms, we were able to demonstrate the ability to accurately measure, and continuously monitor CT acquisition quality. Future algorithms will build on this research to automatically characterize and correct for image acquisition variability.

RSNA 2011 was an exciting event, notably because of the connections we made or cultivated while in Chicago, and we’re looking forward to seeing you all there again next year!

Leave a Reply