RSNA 2015

Abstract Archives of the RSNA, 2015


QRR005

Interoperable communication of quantitative image analysis results using DICOM standard

All Day Location: QIRR, Learning Center



Andriy Fedorov, PhD, Boston, MA (Presenter) Nothing to Disclose
Daniel L. Rubin, MD, MS, Palo Alto, CA (Abstract Co-Author) Nothing to Disclose
Jayashree Kalpathy-Cramer, MS, PhD, Charlestown, MA (Abstract Co-Author) Nothing to Disclose
Justin Kirby, Bethesda, MD (Abstract Co-Author) Stockholder, Myriad Genetics, Inc
David A. Clunie, MBBS, Bangor, PA (Abstract Co-Author) Owner, PixelMed Publishing LLC
Michael Onken, Oldenburg, Germany (Abstract Co-Author) CEO, Open Connections GmbH;
David Flade, DIPLENG, Feldkirchen, Germany (Abstract Co-Author) Employee, BrainLAB AG
Rajesh Venkataraman, PhD, Grass Valley, CA (Abstract Co-Author) Employee, Eigen
Jan Bertling, Stockholm, Sweden (Abstract Co-Author) CEO, Hermes Medical Solutions, Inc
Pattanasak Mongkolwat, PhD, Bangkok, Thailand (Abstract Co-Author) Nothing to Disclose
Steve D. Pieper, PhD, Cambridge, MA (Abstract Co-Author) CEO, Isomics, Inc; Employee, Isomics, Inc; Owner, Isomics, Inc; Research collaboration, Siemens AG; Research collaboration, Novartis AG; Consultant, Wright Medical Technology, Inc; Consultant, New Frontier Medical; Consultant, MeBio; Research collaboration, gigmade;
Ron Kikinis, MD, Boston, MA (Abstract Co-Author) Nothing to Disclose
BACKGROUND

Accurate and unambiguous communication of derived image-related information is of critical for the emerging applications of quantitative imaging. Actively researched areas in quantitative imaging include segmentation of normal organs and pathology, spatial registration across time-points and modalities, and performing measurements. Digital Imaging and Communication in Medicine (DICOM), first released in 1993 is the standard used ubiquitously in the commercial medical devices that are used in clinical radiology, dentistry, cardiology and nuclear medicine imaging systems. It is also increasingly used in non-radiological specialties, particularly for visible light imaging, including histopathology, gastroenterology, dermatology and ophthalmology. DICOM defines both communication interfaces and data formats for images and image-related information (including measurements, ROIs and segmentations). The data formats (object classes) can be used separately (e.g., stored on media) from the communication interfaces. While a variety of DICOM object classes exist for describing such derived image-related information, thus far they found very limited acceptance both in the academic community and among the manufacturers of radiology workstations implementing quantitative image analysis methods. As a result, longitudinal tracking, comparison of methods, and secondary analyses are challenging, while quantitative imaging assessment of patients is difficult or impossible to perform across different platforms. We aim to address these issues by communicating image segmentation results using the object classes defined by the standard.

METHODOLOGY/APPLICATION

In this exhibit we will focus on demonstration of support for communication of image segmentation results. Image segmentation is a common task in quantitative image analysis concerned with partitioning image into regions based on certain application-specific characteristic. DICOM standard defines Segmentation Information Object Definition (IOD) (DICOM SEG) as the image object representing a classification of pixels in one or more referenced images. DICOM SEG is a versatile object that maintains detailed provenance record about the imaged subject and reference imaging data, and provides unambiguous specification of the anatomy being segmented using structured terminology. DICOM SEG also facilitates communication of the information about the segmentation method and software used, as well as information about reader for segmentations performed manually. Support of DICOM SEG has now been implemented in several research and commercial tools, including tools that are free open source, represented by the groups participating in this exhibit. DICOM SEG has also been adopted as the format for communicating algorithm comparison challenges conducted by the NCI Quantitative Imaging Network (QIN), producing a growing publicly available collection of the representative analysis results segmenting publicly available cancer-related image datasets.

DEMONSTRATION STRATEGY

The exhibit will have different components described below: 1. Software demonstration component will provide live demonstrations of free open source publicly available quantitative imaging analysis workstations (specifically, 3D Slicer, ePAD and ClearCanvas AIM), as well as live and/or recorded demonstrations of the capabilities of the commercial products (specifically, BrainLab brain surgery navigation and planning system, Eigen MRI-US fusion planning system and Hermes Medical Solutions workstation) supporting DICOM SEG. 2. Community repositories demonstration component will discuss DICOM SEG segmentation results available on the NCI Cancer Imaging Archive (TCIA). 3. Connectathon component will demonstrate how DICOM SEG objects generated by one participating workstation can be interpreted by another workstation. A collection of sample DICOM SEG objects produced by the participating workstations will be publicly available. 4. Educational component will provide in-depth coverage of the capabilities provided by DICOM SEG and supporting DCMTK toolkit in the form of poster. 5. Tool inventory component will be concerned with creating a publicly accessible catalog of the software tools that support DICOM SEG (software capabilities, sample datasets, relevant screenshots). In addition to the demonstration specific to communication of DICOM SEG objects, the participants will present their related capabilities in integrating DICOM SEG in workflows for reporting quantitative image analysis results using DICOM Structured Reporting and other relevant standard-based objects.

REFERENCES AND PUBLICATIONS

Clunie, D. 2007. "DICOM Structured Reporting and Cancer Clinical Trials Results." Cancer Informatics 4. Libertas Academica:33-56.Fedorov, A, et al. 2012. "3D Slicer as an Image Computing Platform for the Quantitative Imaging Network." Magn Res Imag 30(9):1323-41.Clark, K, et al. 2013. "The Cancer Imaging Archive (TCIA): Maintaining and Operating a Public Information Repository." J. Dig Imag.

Meet-the-Experts Schedule:

Tuesday 12:15pm - 1:15pm Wednesday 12:15pm - 1:15pm Thursday 12:15pm - 1:15pm

Honored Educators

Presenters or authors on this event have been recognized as RSNA Honored Educators for participating in multiple qualifying educational activities. Honored Educators are invested in furthering the profession of radiology by delivering high-quality educational content in their field of study. Learn how you can become an honored educator by visiting the website at: https://www.rsna.org/Honored-Educator-Award/

Daniel L. Rubin, MD, MS - 2012 Honored Educator
Daniel L. Rubin, MD, MS - 2013 Honored Educator