YouTube Facebook LinkedIn Google+ Twitter Xingrss  

May 9, 2003 | ACT scan of the human chest can run in the order of 70 MB. Hospitals and medical research involving imaging such as MRI, X-rays, and PET scans face huge image storage and analytical challenges as digital media replace film.

The picture archiving and communications systems in hospitals face increasing file sizes and new compression methods and standards. Solutions involve putting storage systems in place that interface with other IT systems in the hospital and in the medical research community. This is about not only managing patient data but also using it to propel medical research.

According to Matthew Frosch, a neuropathologist at Massachusetts General Hospital, imaging techniques could become a key diagnostic test for many diseases. Attempts to find and use biomarkers or imaging techniques as diagnostics have been unsuccessful with Alzheimer's disease and Parkinson's disease. For example, it is not clear whether the plaque deposits in Alzheimer brains increase steadily or in a more punctuated way. "Imaging studies may or may not reveal that. So, while image databases are great, they will not solve our basic problems in understanding the disease," Frosch explains.

As faculty coordinator for the Center for Molecular Pathology at the Harvard Center for Neurodegeneration and Repair (HCNR), Frosch is setting up a cross-institutional database between four existing brain banks, which contain tissue from victims of neurodegenerative diseases. These tissue banks are all in the Boston area but are not linked; hence, researchers must approach them individually with a request.

Setting up this database to permit better tissue access for research has turned out to be quite cumbersome. This is surprisingly typical, according to HCNR Director Adrian Ivinson. The center was founded a year ago with a philanthropist's gift. Despite Boston's enviable resources in neuroscience, what has been missing, Ivinson says, are the links between various research efforts to catalyze new collaborations, accelerate the pace of translational research, and make the fruits available to patients.

Even in this networked research world, a lab working on genomics and proteomics may not be aware of colleagues who are staining pathology samples and imaging them while, somewhere else, a person is working on patient records, often handwritten, annotated with lab readouts, and someone else is working on that patient's MRI. "Just those four data sets never come together, except in the note form in a patient's file," Ivinson explains. "So one important, long-term objective at our center is to figure out how to assure that data can be stored and shared in a meaningful way."

The center is also working to develop the infrastructure backbone to allow storage and retrieval of image files such as those from optical microscopes as well as huge MR files.

A similar image-driven project on a national scale is BIRN, the Biomedical Informatics Research Network, which also focuses on brain disorders. Links between a limited number of universities are currently being set up to allow large-scale data exchange. These projects also necessitate a cultural shift. "Individual researchers have images, store them on a CD, and no one else can see them," Ivinson says. "We are not terribly good at sharing our data, and that is something we are facing."

—Vivien Marx

Back to The Sharper Image 

For reprints and/or copyright permission, please contact  Jay Mulhern, (781) 972-1359,