CTSC Imaging Informatics Initiative

Back to CTSC_CrossSiteImagingIdeas

Note: This is a work in progress

Overview

The Harvard Catalyst Medical Imaging Informatics Bench to Bedside (mi2b2) Initiative is preparing to provide services for setting up imaging data management systems (IDMS) to support research and clinical efforts involving imaging for Harvard Catalyst affiliated investigators. Two simultaneous (and potentially conflicting) priorities are to 1) provide optimal, customized hardware and software tools for each investigator/ laboratory and 2) to harmonized these solutions to allow for aggregation of the raw and derived medical imaging data for subsequent sharing beyond the investigator/ laboratory.

XNAT has been identified as a potential software solution for providing the core functionalities of a Harvard Catalyst IDMS. A typical IDMS service engagement will include the following steps:

• Requirement gathering
• Gap analysis
• Infrastructure setup
• Data modeling
• IDMS installation and setup
• Data loading
• Testing and evaluation
• Ongoing data entry, data loading/transfer, and system monitoring and maintenance
• System upgrade

mi2b2 Phases

Planning and Information Gathering

The goals of current phase of the initiative are to:

1. Gather concrete and specific requirements for imaging data transfer and management from a number of high-volume imaging projects led by Harvard Catalyst-affiliated investigators that are generating valuable derived image data with the intention to share that data beyond their own laboratory;
2. Generate a list of overlapping requirements from these specific investigators;
3. Prioritize the projects to determine next steps;
4. Deploy a demo IDMS based on the default XNAT source code;
5. Assess the effort required to extend XNAT code to support the features identified.

Prototype Development

CHB

• Grant Lab

MGH

• Dickerson Lab

Use Cases

(Template

• Mission
  • Description of big picture, goal(s) of project
• Participants
• Data Types
  • Project
  • Imaging Modalities
  • Genetic
• Storage needs
• Workflow
  • Other software to be integrated (e.g., Osirix)
  • other features
• Data transfer
• Data Management)

The use cases presented here are based on on-going informal interviews with various labs and email communications.

Clinical

• Dan Iosifescu, MGH
• Biomarkers for appendicitis
• Mild cognitive impairment

Research

• Ellen Grant, Rudolph Pienaar, CHB Data Management (Step 1); Query Formulation (Step 2); Tractography Workflow (Step 3)
• Stephan Voss, CHB
• Simon Warfield, Neil Weisen, CHB Data Management (Step 1); Query Formulation (Step 2); Analysis Workflow (Step 3)
• Bill Hanlon, DF/HCC
• COPD Gene Project, BWH
• Brad Dickerson, Mike Brickhouse, MGH Data Management (Step 1); Query Formulation (Step 2)
• Mark Anderson, BWH Data Management (Step 1); Query Formulation (Step 2)

Project Prioritization Criteria

Possible criteria include:

• Impact
  • Number of users
  • Duration of the project
  • Amount of data (# of subjects, # of scans ...)

• Success Potential
  • Commitment of PI
  • Feasibility: is it a good fit for using XNAT?
  • Availability of project support at each stage of the effort, including requirement gathering, data collection, system testing, continuing IT and informatics support.

Current Status

Test installations of XNAT (1.4 release candidate 2) were done on Windows XP and Linux in VMware. Minor issues in the setup process needs to be fixed to make successful installation on Windows Vista.

Requirement gathering from several research and clinical projects have been conducted.

XNAT 1.4R2 setup/installation process semi-automated.

XNAT extension prototype built.

Explored ways of extracting and loading additional DICOM fields.

7/23/09 Initial user needs assessments have been performed for three use-cases and initial criteria for measuring success of each use case.

7/23/09 Planned approach for setting up XNAT instance at BWH, and identifying an appropriate use-case in the SPL and Image Guided Therapy project.

7/24/09 Requested some example data from Rudolph Pienaar and Brad Dickerson to begin experiments with organizing & uploading onto xnat instances.

Workflow notebook (under development)

look here

Other Software

In addition, an evaluation of dcm4chee|[1] (a DICOM Implementation in JAVA) is in progress. Simon's group runs an instance of dcm4chee. XNAT's DicomServer and DicomBrowser are based on dcm4che.

The Globus MEDICUS (Medical Imaging and Computing for Unified Information Sharing)|[2] has been proposed to provide an image sharing solution.

Reporting and database management of radiological data for clinical trials offered by http://tumormetrics.org/

Evorad Medical Imaging Workstation from http://www.evorad.com/

List of image visualization and analysis software

Neuroimaging

NITRC Funded by the National Institutes of Health Blueprint for Neuroscience Research, NITRC facilitates finding and comparing structural and functional neuroimaging tools and resources. Collecting and pointing to standardized information about tools, this site can help you find the right structural or functional neuroimaging tool or resource and help you decide whether it can help in your research

Commercial

PACS

Fuji Synapse Used at CHB for MR and CT.

Hermes Used at CHB and DFCI for nuclear medicine

GE Centricity system Used at BIDMC, BWH, MGH and DFCI

Amicas web-based application it uses compression techniques to deliver images to the desktop PC for clinical review. It is used primarily by referring clinicians. Used at MGH

AGFA IMPAX Used at MGH.

Workstations

Toshiba Voxar Works on PC. Used at CHB

Merge efilm Works on PC. Used at CHB and MGH

GE Advantage Used at BIDMC, MGH and CHB for postprocessing of CT and MR data.

TeraRecon AquariusNET server connected with PACS to allow clinicians to analyze and segment volumes from routine desktop computers (rather than dedicated workstations). Used at BIDMC and MGH

Siemens Leonardo Used at BIDMC and MGH for Interventional Angiography & Interventional Neuroradiology

MedX, designed especially for medical imaging researchers and system developers. Works on Linux desktop systems. Used at MGH (neurosurgery)

Vital Vitrea for visualization of CT and MR data. Used at CHB, MGH and BIDMC.

Siemens Syngo for 3D image processing. Used at MGH.

Image management

AccelaRAD Virtual Radiology Network (VRN) enables multi-site radiology groups to perform remote reading across geographic boundaries with greater control over report turnaround times and distribution to clients and referring doctors. Used at CHB

Open source

OsirX OsiriX is at the same time a DICOM PACS workstation for imaging and an image processing software for medical research (radiology and nuclear imaging), functional imaging, 3D imaging, confocal microscopy and molecular imaging. For MAC computers. Used at CHB and MGH

myPACS radiology e-teaching files. Used at CHB

k-PACS DICOM Viewing Software for research only. Used at CHB

FSL is a comprehensive library of analysis tools for fMRI, MRI and DTI brain imaging data. Used at CHB, MGH and BWH for research.

MIPAV (Medical Image Processing, Analysis, and Visualization) application enables quantitative analysis and visualization of medical images of numerous modalities such as PET, MRI, CT, or microscopy. Used at BWH

ConQuest Radiology Micro PACS. Used as the Neuro OR desktop micro PACs server at MGH.

SPM has been designed for the analysis of brain imaging data sequences. The sequences can be a series of images from different cohorts, or time-series from the same subject. The current release is designed for the analysis of fMRI, PET, SPECT, EEG and MEG. Works on PC, MAC, Linux. Used at CHB and MGH (training offered at Martinos center)

Research software from Children's Hospital

STAPLE Simultaneous Truth and Performance Level Estimation. Algorithm for assessing a collection of segmentations of an image.

CRkit (Computational Radiology Kit) for automatic and interactive segmentation, registration and visualization.

Research software from Brigham and Women's Hospital

3D Slicer is a multi-platform, free open source software (FOSS) for visualization and image computing.

Research software from Massachusetts General Hospital

Freesursfer is a set of automated and semi-automated tools for the segmentation, reconstruction, and processing of surface-models of the human cerebral cortex. Used for MRI data.