Other Support

Introduction (April 2005)

NA-MIC, the National Alliance for Medical Image Computing, is a National Center for Biomedical Computing and is funded through the NIH roadmap initiative . The goal of the roadmap initiative is to develop trans-NIH activities that are beyond the scope of a single NIH institute or center. Although NIBIB is the institute managing NA-MIC, the management team is trans-NIH and the funds come from the office of the director.

NA-MIC exists within an active research environment, with multiple collaborators and multiple sites. This page provides an overview of the major funded research efforts of the NA-MIC participants, and the relationship of this funding to NA-MIC. Of note here, in addition to NAC and BIRN, many of the Core and Site PIs participating in NA-MIC also have major research activities funded by NIH, by non-NIH federal, and by private sources. Below we outline the goals of NA-MIC, NAC, and BIRN to provide examples of the relationship among these grants, and to highlight the factors that clearly differentiate them as separate entities. Finally, as with most research collaborations, the participating researchers in NA-MIC are involved in multiple research studies funded through multiple funding mechanisms, and we are pleased that we are able to bring the expertise and strength of such successful scientists to the NA-MIC goals and objectives.

Executive Summary of NA-MIC Interactions with NAC and BIRN

Note: A one picture version of this summary can be downloaded here: pdf and ppt. A one-page text version of this summary can be downloaded here pdf.

• NA-MIC
  • National Level Science, National Level Outreach
    • Less than 15% of direct funding stays at BWH
    • Large majority of investigators working at sites other than BWH
  • Compatible with commonly available equipment and resources
  • Core infrastructure that is broadly deployable
  • Focus on tool and algorithm development (evolving Cores 1 and 2)
  • Platform for national effort to develop image computing software engineering framework

• BIRN
  • National Level Science, National Level Outreach
    • Multi-site collaborative clinical research
    • Focus on image calibration, high-speed networking, grid computing, shared databases and large-scale image repositories
  • Adapts and applies algorithms and application software from other efforts
  • Platform for NA-MIC

• NAC
  • Local Level Science, National Level Outreach
    • Close interactions of Local BWH expertise
    • Leverage Unique BWH environment and infrastructure
    • Custom software
  • Develop "pioneering solutions"
  • Focus on science questions driven by diverse independent collaboration
  • Precursor for some NA-MIC development projects

The NA-MIC, NAC, and BIRN efforts form a mutually beneficial constellation of research initiatives, each with a distinct area of focus and set of priorities, but sharing a common development vision. There are natural interfaces between these projects where interoperability is necessary, but in reality each project has unique deliverables that rely and build upon components provided by the other efforts. In particular, the Neuroimage Analysis Center, as a P41 BTR resource center, brings together a set of collaborators who work together closely, primarily at a local institutional setting (BWH) in order to develop pioneering solutions and to disseminate the results nationally, as well as to make the equipment and related facilities available to the wider community. In contrast, the BIRN and NA-MIC efforts are national consortia aimed at creating solutions that will be workable across a range of institutional settings and different situations.

Thus the differences in scope between NAC and BIRN, and NA-MIC, allow NAC to push the envelope in areas where it can draw on local expertise and the unique environment at BWH. For example, the diffusion tensor based white matter neuroimaging efforts within NAC make use of custom and rapidly evolving MRI pulse sequences not yet widely available outside of the BWH environment. Accordingly, the analysis of these images relies on custom software optimized for the computer infrastructure available at BWH. NAC is, however, dedicated to making these unique resources available through its outreach activities, though this is not the exclusive focus of the research efforts. In contrast, a focus on outreach activities at the national level must work within the constraints of commonly available equipment and resources in their efforts to develop an infrastructure that is deployable to all sites. The fruitful interplay of pioneering local efforts, with widely reproducible and robust national efforts, informed by the best practices at multiple sites, thus provides a significant strength to the research agenda.

In addition to the differences in scope that differentiate NA-MIC and BIRN from the NAC, the NA-MIC and BIRN national efforts themselves have very different scientific targets. The BIRN efforts are fundamentally infrastructure development projects to enable multi-site collaborative clinical research. In particular, the BIRN testbed efforts are heavily focused on multi-site image calibration, nationwide high speed networking, grid computing, and shared databases and large-scale image repositories. Key technology components from several collaborating P41 sites have been adopted, and extended where needed, by the BIRN testbeds in order to prove the efficacy of the infrastructure. Importantly however, the BIRN efforts do not directly fund the development of new algorithms or the corresponding application software methodology -- they only support the adaptation and utilization of algorithms and software within the BIRN testbeds. In contrast, NA-MIC was specifically conceived to address the need for fundamentally new computer science approaches. In one sense, NA-MIC can be seen as an effort to develop the next generation of image computing algorithms that exist in a complementary relationship to, and build upon, the BIRN infrastructure. This new category of algorithms and software engineering practice relies on the well-curated data repositories and abundant grid computing resources provided by BIRN to address the large-scale population studies needed to extract statistically significant findings in measurements of the subtle effects of disease and treatment response. Thus in the burgeoning field of biomedical computing, it is vital to pursue several distinct research agendas -- such as those followed by NAC, BIRN, and NA-MIC in accordance with the mandates of the NIH programs they serve -- while cooridnating the efforts so that the scientific results work well together and avoid redundancy.

Examples to Illustrate Research Relationships between NAMIC, NAC, and BIRN

As an example for the relation between NAC and NA-MIC, we briefly discuss the efforts in DT-MRI postprocessing. In the DT-MRI core within the NAC, the focus is on developing novel filtering, segmentation, and tractography methods for tensor data analysis. During the past several years the team has been working to filter the data using tensor valued Markov Random Fields and analysis of tracts using Laplacian Eigenmaps. The emphasis here is on techniques that have immediate application to clinical DT-MRI data using 3D Slicer.

This sets this work apart from work performed in NAMIC, where currently the algorithmic cores are working on more theoretical approaches to both tensor and non-tensor diffusion data. One important focus within NAMIC is to derive new statistical methods for quantitative measurements (UNC, Utah) and also to investigate non-Gaussian models such as diffusion Q-Ball imaging (MGH).

The 3D Slicer software itself is used as a component of all three projects, but has very different uses driven by the needs of each project. NAC continues to develop the core functionality of Slicer for use by the broader scientific community and has a mandate to disseminate these results. It has been successful in coordinating and sharing image analysis software for image analysis software developers at BWH and nearby with solutions customized for the local environment. Dissemination, on the whole, has been successful with several thousand registered software downloads. But within NAC this effort is limited by the the number of analysis tools that can be incorporated and the ability to test the software in a wide variety of end-user environments. Slicer is built on ITK/VTK, which is a growing industry standard but is not yet widely accepted and requires significant effort to train developers. NA-MIC provides the support to integrate and disseminate a much larger portfolio of medical image analysis software contributed by developers far beyond the BWH site. The emphasis and funding for robust software engineering, training and dissemination is unique to NA-MIC and ensures that like the NAC software tools, those from other expert labs are widely available to the scientific community within a compatible software engineering framework. Slicer serves as the visualization platform for the BIRN project, for which specific modules are adapted to meet the needs of BIRN researchers. For example, significant developments have been made to support interoperability with the FreeSurfer software from MGH and the LDDMM software from JHU, both critical components of the Morphometry BIRN testbed.

Background Information on NIH Missions for the NCBC, BTR, and BIRN

• Biotechnology Resources (P41)

From NCRR's Biomedical Technology Guidelines

"These BT Centers provide state-of-the-art experimental and computational resources to a wide range of biomedical researchers, particularly those supported by NIH."

• National Centers for Biomedical Computing (U54)

From NIH's description of the NCBC program

"The centers will create innovative software programs and other tools that enable the biomedical community to integrate, analyze, model, simulate, and share data on human health and disease."

• Biomedical Informatics Research Network (BIRN)

From NCRR's description of BIRN goals

"The initial goal is to address the needs of biomedical investigators across the country to effectively share and mine data in a site-independent manner for both basic and clinical research."

FAQ

1. Like NA-MIC, NAC also has driving biological problems?

The short answer is yes, but they are called collaborations. The NCBC's are structured in a way that is very similar to the NCRR biotechnology resources. Both have technical development efforts which are responsive to collaborators' needs. NAC, however, is a national resource center and uses neurosurgery, neurologic imaging and schizophrenia as collaborator projects. In contrast, NA-MIC is a U54 and uses several different schizophrenia projects as driving biological problems. Each of these projects have distinctive goals and questions to be addressed.

2. How is the "pioneering" part different in NA-MIC and NAC, i.e., software vs. equipment or different types of software?

Both NA-MIC and NAC work on the general topic of medical image analysis, but there is no relevant overlap of the scientific personnel in the technical development areas. The specific algorithms that are being researched are very different. Accordingly, the specifics of the core algorithmic science are different.

3. How is the tool and algorithm development in NA-MIC different from pioneering efforts and custom solutions in NAC?

4. How is the software development in NA-MIC different from the developments in NAC.

NA-MIC and NAC both use the VTK and ITK toolkits and Slicer as the application delivery platform. NAC software development focuses on the specific NAC applications of MRI neuroimaging. The NA-MIC Engineering Core looks for generlalizations and broadens the impact of algorithms generated by Core 1. Iin addition, Core 2 defines software practices and processes for the Center.

For NA-MIC, the goal is to base the work on a shared engineering infrastructure to enable maximal leverage of the software foundation. NAC aims at enabling local scientists to make the results of their research available on a national scale. NA-MIC is a national alliance of algorithm and software researchers, catering to a national audience.

Major support of the participants in 2004

Each entry represents support of over 150k direct per year with leadership function.

NA-MIC is a truly national alliance. Each of the participating groups has a substantial portfolio of local and non-local activities. NA-MIC both leverages and enables these resources on multiple levels. The previous section discusses this relation in more detail with two examples.

The funding inside NA-MIC is distributed in compliance with the RFA: 50% of the NA-MIC direct funding is distributed in roughly equal parts among the participants of the Algorithm and Engineering cores. 25% is split among the DBP's and the remainder is split among the service, dissemination, training, and leadership cores.

Role in NA-MIC	PI	Grant	Grant Number	Agency
Leadership
BWH	Kikinis	Neuroimaging Analysis Center	P41 RR013218	NCRR
		MR Guided Therapy	PO1 CA67165	NCI
		Virtual Soldier Phase 1	A02-600124005	DARPA
		Image Guided Technology	n/a	CIMIT
		Morphometry BIRN	U24 RR021382	NCRR
BWH	Wong	Bioinformatics Research Center	n/a	HCNR
		Time-lapse Microscopy Cellular Analysis	5R01 LM008696	NLM
Algorithms
MIT	Grimson	ERC Center for Medical Robotics and Computer-Assisted Interventional Systems and Technology	8810-27499	NSF
MGH	DKennedy	Anatomic Morphologic Analysis of MR Brain Images	5R01 NS34189	NINDS
		Neuroanatomy of Adult ADHD	5R01 MH62152	NIMH
		Affective & Conative Changes in Alcoholism	3R37AA007112-1551	BU
		Validating High-Field Functional & Structural MR	??	USArmy
		Neuroimaging Neuroinformatics Training Program	1 K12 MH069281 A01	USArmy
U of Utah	Whitaker	Dynamic MRI for myocardial perfusion and viability	R01 EB000177	NIBIB
		Assembling Visible Neurons for Simulations	10218584	NIMH
		Geometric Surface Processing Tools for Analysis of Biological Data	EIA 0313268	NSF
		Virtual Parts Engineering Research Center	DAAD19-01-1-0013	ARO
		Hybrid Modeling for Geometric Design, Estimation, and Analysis	CCR-0310705	NSF
		High Resolution MR Angiography	2R01 HL048223-11 A1	NHLBI
UNC	Gerig	Medical Image Presentation MIP: Structural Image Analysis and Medical Uses	P01 CA47982-11	NIBIB
		Prospective Studies of the Pathogenesis of Schizophrenia	P50 MH064065-01A1	NIMH
GATech	Tannenbaum	Atmospheric propagation of high energy lasers: modeling, simulation, tracking	0205-G-DB083	AFOSR
		Geometric variational methods for computer vision	DAAD19-02-1-0378	ARO
		Active vision control systems for complex adversarial 3-D environments	F49620-03-1-0401	MURI
		Statistical and variational methods for problems in visual control	F49620-05-0017	AFOSR
Engineering
GE	Lorensen	Virtual Soldier	W81XWH-04-2-0012	DARPA
UCLA	Toga	Computational Anatomy and Multidimensional Modeling/BIRNs	5 P41 RR013642	NCRR
		Computational Biology from Genotype to Phenotype	1 P20 MH065166	NIMH
		A Multidimensional Alzheimer’s Disease Brain Atlas	5 R01 LM005639	NLM
		Characteristics of Perfusion Related Cortical Signals	2 R01 MH52083	NIMH
		A Probabilistic Reference System for the Human Brain	9 P01 EB001955	NIBIB
UCSD	Ellisman	IVEM and Image Analysis Resource (NCMIR)	5P41 RR04050-16	NCRR
		Biomedical Imaging Research Network – Coordinating Center (BIRN-CC)	1R24 RR019701-01	NCRR
		Dynamics of Membrane Organization at Nodes of Ranvier	5R01 NS14718-21	NINDS
		National Partnership for Advanced Computational Infrastructure (NPACI)	ASC 97-5249	NSF
		3D Cell-Centered Neuronal Database	9R01 DA016602-07	NIDA
Kitware	Schroeder	Insight Toolkit (ITK): Image Processing Tools for the Visible Human Project	N01-LM-9-3532	NLM
		Visualizing Arbitrary Basis Functions for Advanced Engineering Analysis and Simulation (SBIR)	DMI-0128453	NSF
		Simulation-Based Medical Planning for Cardiovascular Disease	0205741	NSF
		Scalable Grid Technologies for Visualization Services (SBIR)	DE-FG02-03ER83692	DOE
		Software Toolkit for Image Guided Surgery (STTR)	2 R42 EB000374-02	NBIB
		System Technology for the Insight Toolkit ITK	NLM RFP 04-101/VMS	NLM
		Micro-to-Macro Multimodality Atlas Formation (STTR)	1 R41 EB004737-01	NBIB
		3D Widgets for Segmentation and Registration	NLM 04-177/CYC	NLM
		The Insight Journal: an Open Access Repository for Technical Papers related to the Insight Toolkit	NLM 04-177/CYC	NLM
		The Visual Database: Portable, XML-Based Middleware For Media Representation, Interaction and Exchange (SBIR)	DMI-0450513	NSF
Isomics	Pieper	Neuroimaging Analysis Center	P41 RR13218	NCRR
		Virtual Soldier Phase I	BAA 03-02	DARPA
		Morphometry BIRN	U24 RR021382	NCRR
Driving Biological Projects
VA/Harvard	Shenton	Computerized Image Analyses of MR Scans in Schizophrenia	2R01 MH50740	NIMH
		NeuroImaging Studies of Schizophrenia	n/a	VA
		Neurophysiological Studies of Schizophrenia	RO1 MH40799-09	NIMH
		Clinical Symptoms & Brain Abnormalities in Schizophrenia	K05 MH70047	NIMH
		MR Brain Diffusion Tensor Imaging in Schizophrenia	n/a	VA
UCI	Potkin	Transdisciplinary Imaging Genetics Center	RR20837	PHS
		Genetic and Neuropathological Abnormalities in Neuropsychiatric and Neurological Disorders	PNDRF28409	Pretzker Foundation
		Functional Imaging Research in Schizophrenia Testbed: Biomedical Informatics Research Network (BIRN)	M01-RR000827	NCRR
		Neuropathological and Genetic Abnormalities in Depression	MH60398	NIMH
		Morphometry Biomedical Informatics Research Network (mBIRN)	MGH36106	NCRR
Dartmouth	Saykin	Neural Mechanisms of Chemotherapy-Induced Cognitive Disorder	R01 CA101318	NCI
		Memory Circuitry in Mild Cognitive Impairment and Early Alzheimer's Disease	R01 AG19771	NIA
		Cognitive Effects of Cancer Chemotherapy	R01 CA87845-03	NCI
		Dopaminergic modulation of working memory in TBI: An fMRI study	R01 NS40472	NINDS
		Catecholaminergic Modulation of Working Memory in TBI	R01 H133G000136	NIDRR
		IDM: Data Management of Protected Information for Data Sharing and Collaboration	00308229	NSF
		A System for Data Integration and Pattern Discovery in Multimodal, Spatio-temporal Data	0312629	NSF
U of Toronto	JKennedy	Genetics of Childhood Onset Depression: Relationship to Psychological Measures	P01 MH056193-08	NIH
		Supplement to Risk Factors in Childhood-Onset Depression	P01 MH56193-03S1	NIH
Service
Kitware	Schroeder	See above
Training
MGH	Gollub	Morphometry BIRN	R24 RR 012382	NCRR
		MGH General Clinical Research Center	5MO1-RR001066	NCRR
		MIND Institute	n/a	DoD
Dissemination
Isomics	Pieper	See above