slide1 l.
Skip this Video
Loading SlideShow in 5 Seconds..
Summary of Adoption Activities Planning PowerPoint Presentation
Download Presentation
Summary of Adoption Activities Planning

Loading in 2 Seconds...

play fullscreen
1 / 30

Summary of Adoption Activities Planning - PowerPoint PPT Presentation

  • Uploaded on

Summary of Adoption Activities Planning. Joel Saltz, MD, PhD Department of Biomedical Informatics Ohio State University Medical Center, Columbus, OH. Goals of Imaging Related Cooperative Group Efforts.

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Summary of Adoption Activities Planning' - boyd

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
summary of adoption activities planning

Summary of Adoption Activities Planning

Joel Saltz, MD, PhD

Department of Biomedical Informatics

Ohio State University Medical Center, Columbus, OH

goals of imaging related cooperative group efforts
Goals of Imaging Related Cooperative Group Efforts
  • Groups actively engaged in Imaging Workspace collaborations: ATC, QARC, CALGB, ACRIN, RTOG, NTROI, Children’s Oncology Group
  • Goal: Maximize reproducibility in staging, grading of cancer
  • Approach: Grid based “central” review, algorithms
  • Goal: Better characterize disease to target treatment
  • Approach: Molecular imaging, image analysis
  • Goal: Maximize precision in targeting tumors
  • Approach: Image guided radiation treatment, adaptive radiotherapy, adaptive image guided surgery
  • Goal: Imaging as a biomarker
  • Approach: Algorithms to reproducibly quantify changes in response to treatment
maximize reproducibility
Maximize Reproducibility

Multi-modality Reconstructions of Therizinosaurus

ivi translational research efforts
IVI Translational Research Efforts
  • Coordinated workspace support for QARC/ATC/CALGB/ACRIN central review
  • Grid enabling NCIA, leverage NCIA infrastructure for cooperative group data management
  • Support for grid enabled CERR and integration of grid enabled CERR into ATC, QARC protocol managment activities and into the QARC/ATC/CALGB/ACRIN central review framework
  • Pathology central review support for NLST
  • Pathology image quantification for Children’s Oncology Group (H&E)
  • Grid based Breast Cancer Pathology TMA analysis (with David Foran)
  • gACRIN support
cagrid enabled cerr
caGrid Enabled CERR
  • Access images, store review results at ATC, QARC, ACRIN, CALGB image archives
  • Matlab code can run locally or on remote clusters
  • Markup annotation of images via caBIG AIM standard
gridimage cagrid integration
gridIMAGEcaGrid integration

Leverages core caGrid services/tools

Introduce, caDSR Service, caGrid Data Service, Index Service, Authentication Service, dynamic WSRF resources

Leverages In Vivo Imaging Core Middleware

DICOM interoperability and Bulk Data Transport via GridFTP

image quantification neuroblastoma with shimada children s oncology group
Image Quantification – Neuroblastoma (with Shimada, Children’s Oncology Group)
  • International Neuroblastoma Prognosis Classification System developed by Shimada et al., classifies the disease into various prognostic groups in terms of different pathologic features
  • Service-based infrastructure
    • Multiple, geographically distributed scientists and developers access a common image data repository
    • Shared code repository allowing reusability of the developed codes
    • Remote job execution
    • Remote image visualization
  • AVT Prototype: Multi-processor backend
    • Fast parallel processing of images
    • Specifically designed for very large-scale image processing
    • Pipelined processing capabilities
neuroblastoma grid service
Neuroblastoma Grid Service
  • The service is developed
    • Based on the caGrid 1.0 middleware
    • Using Introduce service development toolkit
      • Strongly-typed interfaces
  • Provided operations on images/algorithms
    • Query
      • CQL (caGrid Query Language)
    • Retrieve/Upload/View
      • Bulk data transfer
      • GridFTP
    • Execute
      • Out-of-core virtual microscope
parallel processing infrastructure
Parallel processing infrastructure

Processor 1

Processor N


Parallel Classification

Classification map

Assign classification labels

Label 1



Label 2

Whole-slide image

Image tiles (40X magnification)



NLST: caMicroscope Server Architecture

  • Image Data Services that store images in Aperio SVS format
  • Analytical service that pulls images from the image server and executes MATLAB programs on a cluster
  • All services are caGrid 1.0, use gridFTP for high performance bulk data transfer
  • The Client allows multi-resolution browsing of the images, invoke MATLAB programs and examine the MATLAB output images
workspace aim 1 multi cooperative group grid platform
Workspace Aim 1: Multi-Cooperative Group Grid Platform
  • Target three protocols
  • Develop Prototype Grid facing interfaces to ATC, ITC, QARC, CALGB, ACRIN data stores
    • NCIA likely to play major role
  • Develop XIP Based VIEW Review Client and integrate Grid Enabled CERR (thin and thick client) to support Radiation treatment planning
  • Develop caBIG data models to encompass all image, correlative data and all AIM metadata associated with targeted protocols
  • Develop workflow, semantic/federated query engine, harden annotation server, incorporate DICOM worklist support
  • Develop Prototype Harmonized Security and Auditing Infrastructure
  • Employ workflow engine to support QA and central review tasks required by the targeted protocols
workspace aim 2
Workspace Aim 2
  • Extending IVI imaging technology to Pathology, deployment
  • End to end Pathology central review, image quantification use case analogous to Aim 2
  • NLST engagement ongoing; short term potential for CALGB, Children’s Oncology Group
aim 3 integrated analysis of image molecular clinical data
Aim 3: Integrated Analysis of Image, Molecular, Clinical Data
  • Emerging approach that IVI needs to engage
  • Several IVI SMEs actually are doing this but it is not yet well integrated into workspace activities
  • Need to engage ICBP, TCGA, SPOREs etc
summary of end to end soup to nuts cradle to grave
Summary of End to End, Soup to Nuts, Cradle to Grave …
  • Rapidly develop a prototype infrastructure that demonstrates that IVI Workspace technology can support secure cooperative group database interoperability and central review.
  • Motivate this work by three protocols RTOG protocol 0522, COG AHOD protocol 0031 and CALGB protocol 80302 (?)
  • Exemplary protocol studies will be distributed between QARC, ACRIN and CALGB databases.
  • XIP based review client will be developed and integrated with the Grid Enabled CERR client. Legacy QARC, ACRIN and CALGB client software will also be able to access all studies.
  • Software developed through this effort will be used to support the three targeted protocols and will be a first step towards the development of an integrated VIEW/ATC software framework.
develop workflow engine to support qa and central review tasks required by the targeted protocols
Develop workflow engine to support QA and central review tasks required by the targeted protocols

The workflow engine needs to coordinate any automated QA to check to make sure that appropriate files have been uploaded. The workflow engine also coordinates the central review process.

This would be likely to include

  • Finding out which reviewers need to be consulted,
  • Obtaining any correlative data or past studies that need to be included in the review process,
  • Notifying reviewers that their services are needed and coordinating with reviewer client software to manage image display and markup and
  • Inserting reviewer annotations and markups into the AIM data service. Once all reviewers have completed their evaluation, the workflow would provide images, correlative data and all markup information to an adjudicator who would inspect and evaluate reviewer markups and annotations.
develop workflows queries and analyses to support off line data analyses
Develop workflows, queries and analyses to support off-line data analyses
  • Data analyses:
  • RTOG 0515: Determine the impact of PET/CT fusion for each patient by comparing gross tumor volume (GTV) contours and 3DCRT treatment plans using two separate data sets (PET/CT and CT only)
    • Determine the impact of PET on the following endpoints: GTV (cm3), number of involved nodes, location of involved nodes, and dosimetric measures of normal tissue toxicity (mean lung dose, V20, and mean esophageal dose).
  • RTOG 0522: Assess the role of FDG-PET/CT scans in determining the overall clinical outcome and the need for nodal dissection. Test hypothesis that:
      • pre-treatment SUVmax>mediun predicts poor clinical outcome
      • negative post-treatment PET in patients with N2-3 disease predicts for high pathologic complete response rate in the neck
      • negative post-treatment PET in patients with N2-3 disease predicts for low overall nodal replace rate
qc data analyses
QC Data Analyses
  • Support credentialing, quantify interobserver/inter-site variability in
    • RECIST (or other) tumor measurements
    • Radiation treatment plans
    • Pathology diagnosis
  • Develop a common set of caGrid based APIs for data retrieval and submission and then to use pluggable modules based on this API to support both XIP clients and legacy clients.
  • As part of this effort, we will identify or create new data models to describe the different types of data managed in the protocol (e.g. DICOM images, radiation treatment plans as DICOM-RT or CERR, clinical, outcome, molecular and pathology data).
develop prototype grid facing interfaces to ncia itc acrin qarc calgb
Develop Prototype Grid facing interfaces to NCIA, ITC, ACRIN, QARC, CALGB
  • The grid facing interfaces will use a common language for query and a standard API for data retrieval and submission. The use of a standardized API will allow pluggable modules to be developed for the various clients that the cooperative groups have developed. These modules will provide access to the different cooperative group data management systems via the grid interfaces. These modules will be developed for clients such as ACRIN’s TRIAD, Dicommunicator and Dicommunicator.NET from QARC, CERR from RTOG.
develop prototype xip based review client
Develop Prototype XIP Based Review Client
  • Specific Instance of David Channon Project (?)
Develop caBIG data models to encompass all image, correlative data and all AIM metadata associated with targeted protocols
  • Discussed yesterday (!)
develop ivi middleware support needed for stn consortium effort
Develop IVI Middleware Support needed for StN Consortium Effort
  • Workflow: New infrastructure needed includes:

Support for humans in workflow loop

Use of identifiers to refer to images and correlative data (DICOM has this but more work is needed in caGrid to support identifiers)

Support for rules (joint between Middleware and AIM teams)

  • DICOM: DICOM Normalized Messaging support and DICOM worklist support in IVI Middleware.
  • Annotation server: We will complete development of annotation server and use the annotation server for storage, query and retrieval of protocol AIM data.
  • Federated Query Server: We will develop a federating query server which will allow reviewers to retrieve the annotations, and all associated image and non-image correlative data.
develop prototype harmonized view security and auditing infrastructure
Develop Prototype Harmonized VIEW Security and Auditing Infrastructure
  • Pluggable modules will be used to support a harmonized security infrastructure. The grid-facing interfaces of the databases will include components to provide role-based access control, auditing of transactions and provenance tracking. A secure virtual organization will be created and trust relationships will be established between the identity providers of the individual investigators.