New Center for Brain Imaging Science
From the Radiology Department Newsletter, Editor: Mary McAllister

Dr. Susumu Mori and Dr. Mike Miller

How do you improve on an already-world-class imaging facility?
With the MRI Division, the FM Kirby Center, the NMR Service Center, and the BRB Small Animal Service Center, Hopkins already has one of the best imaging facilities, which produces good, high quality images.

“We have more than kept pace with the amazing advancements in MRI and PET over the last 15 years,” said Dr. Susumu Mori, Professor of Radiology and head of the newly established Center for Brain Imaging Science. He went on to explain that “the bottleneck was in moving from image acquisition to image analysis.”

Dr. Mori wanted to address that critical need, and when the Brain Science Institute, established with donor funding, put out requests for grant applications to establish a core-like facility that would serve the entire JHU community, Dr. Mori saw the opportunity.

“Dr. Mike Miller of BME and Dr. Marilyn Albert in Neurology also had potential grant applications for this program, and I thought it would be a stronger proposal with their input, and to combine all three applications.” He also incorporated a PET complement.

“Our proposals [Dr. Miller’s and Dr. Mori’s] focused on image and analysis, and Dr. Albert’s focused on education—an important component that we didn’t have,” said Dr. Mori. He went on to explain that Dr. Albert’s educational focus will provide instruction about how to use state-of-the-art image analysis software, and spur the dissemination of software and technology.

Dr. Mori said that the new Center is part of an effort to change the way imaging is obtained and analyzed. “Imaging is still a matter of human, subjective judgment, and qualitative evaluation. What we would like to do is make imaging quantitative, objective, and automated.”
At the core of the center is the electronic brain atlas and image transformation technology. The atlas stores information about brain anatomy, including normal shape and normal variability. The transformational technology transforms the atlas to an individual’s brain shape. In the case of brain pathology, this will enable quantitative comparisons with the normal population.

The plan is to establish a service with four main cores: 1) a protocol development core that will offer help in developing imaging protocols specific to the problem to be studied; 2) an image-analysis core that will be fee-based, and will convert imaging to numbers that capture the characteristics of a specific brain; 3) a grant support core, which will offer financial support and help in grant writing, such as the Preliminary Data and Methods sections; and 4) an education core, that will offer instruction in how to use cutting edge image processing software.

“We want to reinforce qualitative judgments with numbers, much like a blood test gives a +/- normal range. For example, when we assess the impact of stroke infarction on brain functions, we could determine which area and how certain areas of the brain are affected, and help in developing a more accurate prognosis. Without the use of quantitative parameters (numbers), we can’t create an XY plot, where X represents functional loss and Y represents a particular image feature. In the dozens of numbers generated from the brain, one or more will correlate with function, and we will be able to determine who would benefit the most from rehabilitation or other types of treatment,” Dr. Mori said.

Funding from the Brain Science Institute is for two years, and Dr. Mori is confident that the Center for Brain Imaging Science will be a vital, self-sufficient part of the Hopkins community by then.