In the past year, we put a lot of effort into generalizing DICOM conversion tools and preparing dcmqi (DICOM for Quantitative Imaging). dcmqi is a collection of libraries, commmand line tools, web applications, and other goodies, that aim to help in converting data stored using research formats into DICOM representation. At the moment, we have functional tools that can be used to read and write DICOM Segmentation and Parametric Map objects, and convert measurements into DICOM Structured Reporting documents that follow TID1500 reporting template.
Over the last month, dcmqi has been put to test as some of the participants of DICOM4QI use it as part of their platforms to support DICOM interoperability. While we definitely have a lot of work to do, we are happy with the current status of dcmqi. (N.B.: we are NOT happy with the current status of dcmqi documentation, but we will definitely improve it!)
We have integrated dcmqi with 3D Slicer, our main platform for delivering end-user functionality. To support end-user workflows, we have re-implemented from ground up the QuantitativeReporting extension of 3D Slicer. The updated extension now allows you to segment DICOM data, compute segmentation-based measurements, save the results in DICOM and support loading of the measurements results stored using DICOM SR back into Slicer. See the screenshot below for a snapshot of the user interface!
Michael Onken and Jörg Riesmeier led the development of new functionality in the DICOM Toolkit (DCMTK). New this year are helper classes to support reading and writing of DICOM Parametric Maps, bug fixes to the support of Segmentation objects, and new classes for handling TID1500 structured reports. All of those developments are available in the latest snapshot release of the toolkit, and are now integrated both with 3D Slicer and dcmqi.
This year we published the paper that summarizes a lot of our efforts in the past 2-3 years. At RSNA 2015 we shared a preprint of this paper, and now we are proud to report the work is completed and peer-reviewed paper is published, accompanied by the open-source tools and a publicly available dataset on TCIA. The quantitative analysis tools discussed in that paper are publicly available as 3D Slicer extensions - learn more in the videos prepared by the QIICR Iowa team.
Funded by an administrative supplement to the Brigham and Women`s Hospital and Stony Brook University (led by Joel Saltz) ITCR teams, we have succeeded in integrating OpenCV library into 3D Slicer, and releasing SlicerPathology extension. This new extension is designed to assist with the tasks of segmenting digital pathology images, and is integrated with caMicroscope for convenient access to the data. OpenCV is a library aimed at real-time computer vision processing, which is now easily accessible to all of the 3D Slicer developers in the SlicerOpenCV extension. We are grateful to the ITCR program for supporting this development!
Check out the NA-MIC wiki page for more information on 3D Slicer related activities at RSNA 2016 including the 3D Slicer booth in the Quantitative Image Reading Room and courses using 3D Slicer and other open source tools for 3D printing applications.
We look forward to meeting you at RSNA 2016!
Andrey Fedorov OUTREACH