Neuroinformatics: Exploring the Human Brain

A human brain mapping system proposed by MITRE scientists promises to revolutionize how neuroscience research is conducted

January 2002

Exploring the Human Brain

A human brain mapping system proposed by MITRE scientists promises to revolutionize how neuroscience research is conducted, speed the discovery of new drugs, and improve clinical decision-making for medical and psychiatric disorders over the next decade.

The effort to create a human brain mapping system is also expected to establish MITRE as an important resource for neuroinformatics and bioinformatics expertise worldwide.

Neuroscience researchers are accumulating vast amounts of brain mapping data, but there is no central repository for storing or sharing it with other investigators, according to Lead Artificial Intelligence Engineer Jordan Feidler. MITRE has just received funding from the National Institute of Mental Health (NIMH) and the National Science Foundation (NSF) to develop a worldwide information infrastructure that researchers can use to share, analyze, and explore this data.

Neuroinformatics Defined

To understand how the brain works, researchers and clinicians look at anatomical structure, function, and neurochemical patterns. Although a structural map of the brain laid out nearly 100 years ago is still widely used today, the advent of non-invasive techniques, such as magnetic resonance imaging (MRI) and positron emission tomography (PET), has helped to advance the diagnosis and treatment of brain dysfunction and related conditions.

Taken together, structural and functional information help to form a "map" of the human brain for investigators to study.

Neuroinformatics seeks to create an information infrastructure for sharing and exploring large quantities of this data.

ICBM Leading the Way

Leading the effort to compile this information is the International Consortium of Brain Mapping (ICBM). Formed in 1993 with a program grant from the NIMH, this worldwide group composed of 13 academic institutions plans to make brain mapping data freely available to researchers and clinicians–and that’s where MITRE comes in.

"What our team will focus on during the first five years of this effort is structural MRI data, but the overall vision is that the system will include functional data as well," said Feidler. "ICBM is already collecting this type of data now.

Two Layer Architecture Proposed

According to Lead Database Scientist Dr. Kenneth Smith, the proposed system would have two levels to organize the data and make it accessible. "The lower layer will contain a digital library, whereby users can upload and download three-dimensional brain images. This level will also contain descriptive data on a given study–everything from the time of day the images were collected and what scanner settings were used to the patient’s age, sex, and test scores.

"The upper level, which we call the ‘data warehouse,’ will include a content-based retrieval mechanism and will allow investigators to use the images to create average representatives for selected demographic groups, called atlases. This gets to the core of the ICBM’s effort," he added, "which is to create a reference system that allows researchers to account for normal variability between subjects."

Investigators will access the digital library through a simple browser interface. A specialized access control module will ensure that constraints based on subject privacy and the data-sharing preferences of the principal investigators are met.

"When we share this vision with the neuroscience community, they are very excited," noted Smith, who presented the proposed system in June at the Human Brain Mapping Conference in Brighton, England. "They’re happy to see someone with the expertise and experience in information science that MITRE has crossing the barrier into their world "

The Challenges Ahead

MITRE faces a number of challenges in building the system. "Creating the digital library will be challenging," continued Smith. "However, creating the data warehouse, with its notion of generating probabilistic brain atlases dynamically, really is ‘going where no man has gone before.’ There are so many unexplored variables."

Moreover, according to Feidler, clinicians and researchers differ in their use of terminology and interpretation when entering descriptive data. Lack of a standard language makes it difficult to transfer findings between groups.

"The issue of interoperability for large-scale projects is something in which MITRE is very experienced," said Smith.

"If MITRE’s work helps create a consistent framework for interoperability for the entire human brain mapping community, the importance of this contribution would be enormous." Feidler added that such a framework would make it possible to test hypotheses that involve large quantities of data, because barriers to interoperability would be removed. Both scientists also pointed to the "social" challenges–spanning the divide between the neuroscience and information science communities and maintaining effective communication between both groups.

Important Benefits Seen

Feidler and Smith agreed that neuroinformatics will likely speed certain types of basic research and the discovery of new drugs for conditions like Alzheimer’s disease, Parkinson’s disease, and some psychiatric disorders. "The system has the potential to allow investigators to test their hypotheses on data that’s already been collected," said Feidler. "It will help them crystallize an idea very quickly to determine whether it’s worth pursuing."

Smith pointed to the clinical impact such a system could have. "Let’s say a person is about to have neurosurgery," he said. "With these advances, the neurosurgeon will have access to a probabilistic brain atlas of someone in that patient’s demographic group and similar medical history. Every bit of information that I can get into that surgeon’s hands is important."

Besides the obvious impact on medical outcomes, the system might also be useful one day for educational purposes, perhaps by allowing teachers to download some of the images for study.

"Part of MITRE’s mission is to work in the public interest," said Feidler. "Helping others learn more about the brain and get excited about neuroscience means that we can have an impact on many different levels."

Feidler added that certain components of the system would be useful to a number of MITRE’s traditional sponsors, particularly those managing massive amounts of data that need to be retrieved quickly. "We’ve briefed a number of agencies in the intelligence community, and they have been very enthusiastic about the idea of a content-based retrieval system," he said.

"They especially like the fact that we have an opportunity to build a system using advanced research and technologies without it being limited by legacy systems."

"There’s quite a difference between simply having large amounts of data on a disk and having a whole architecture and information infrastructure enabling users to explore and abstract this kind of data," added Smith. "Such a system would be very useful to many of MITRE’s customers."

Getting Started

Feidler and Smith, and their MITRE team, have been meeting with members of the ICBM in San Antonio, Los Angeles, and Montreal to establish a network of contacts, examine existing information systems, and work out the details of the program’s first year effort.

"It’s an exciting opportunity. Often we feel like we’ve dived into the ocean and been hit by a big wave," described Smith. "We’re learning as we go."

–Michele Gauthier

The MITRE Corporation joins researchers and clinicians around the world, as part of the International Consortium for Brain Mapping, in creating an information infrastructure to share the large quantities of data available on brain structures and functions. Although a structural map of the brain laid out nearly 100 years ago is still widely used today, neuroinformatics seeks to use today's technologies to create a "map" of the human brain.

Solutions That Make a Difference.®
Copyright © 1997-2013, The MITRE Corporation. All rights reserved.
MITRE is a registered trademark of The MITRE Corporation.
Material on this site may be copied and distributed with permission only.

Privacy Policy | Contact Us