Pat Barta

"Stereologic Analysis of Volume Data: 3-D Medical Imaging"

ABSTRACT -- Many imaging techniques used in biology and medicine produce volume data. Volume data are a 3-dimensional generalization of ordinary 2-dimensional digital images.

When dealing with 2-D images, it is convenient to divide the field into two parts: image processing and image analysis. Image processing is the mathematical manipulation of data in images to yield new images, while image analysis is the measurement of various features seen within these images. The methods used for measuring features in 3-D are substantially different from those used for measuring in 2-D.

The goal of this talk is to describe and to illustrate these "stereologic" methods for 3-D analysis. Stereology relates 3-D topological parameters defining a feature to 2-D measurements obtainable on sections through the structure. Stereologic methods were primarily developed by microscopists in conjunction with applied mathematicians working the fields of integral geometry and geometric probability.

Examples described in this talk will be taken from the speaker's research into brain changes in various psychiatric disorders including Down syndrome, schizophrenia, bipolar disorder, HIV, and dementia.

SPEAKER -- Patrick Barta was born in Bryan, Texas, and moved to Baltimore in 1974 to attend the Johns Hopkins University. Since coming to Hopkins, he obtained an undergraduate degree in bioengineering, a joint MD-PhD degree in biomedical engineering, and completed a residency in psychiatry. He is currently an associate professor of psychiatry at the Johns Hopkins Hospital, and divides his time between teaching, patient care, and research into the biological basis of psychiatric disorders.

Colloquium Committee Sponsor: Jim Heaney, Swales Aerospace, 301-902-4531
Next Week: "Do We Need Computers in Schools?", Cliff Stoll
Engineering Colloquium home page: http://ecolloq.gsfc.nasa.gov