Computational neuroscience: Unraveling mysteries of the brain

A three-pound mass of gray tissue housed in a bony encasement has been the focus of our curiosity, awe and wonder for hundreds of years. It generates literature, art, war, love and reason; it is what makes us human. Yet very little is known about how the brain works, human or otherwise.

Scientists are just beginning to unravel how the various parts of the brain function together to allow us to read, walk and respond to stimuli. As the end of the 20th century approaches, the neurosciences are increasingly active, using new approaches and technologies, such as mathematical modeling and MRI scanners, to learn more about the brain and how it works.

In 1995, Philip Ulinski, Professor in Organismal Biology & Anatomy -- along with Terry Regier, Assistant Professor in Psychology, and Jack Cowan, Professor in Mathematics and Neurology, and their colleagues -- began offering a lecture series on how the brain works. The series has since become the core of the University's computational neuroscience program, now in its third year.

A relatively new field, computational neuroscience uses an interdisciplinary approach to address such questions as How does attention arise? and How do neurons form and store memories?

The discipline also employs what Ulinski calls "reverse engineering" -- building robots or computer simulations based on brains of lower animals, and then studying those mechanisms to get a better idea of how the brain works. Robots built for this purpose have been used for a wide variety of tasks, from exploring the ocean floor and cleaning up toxic sites to locating people in collapsed buildings.

The Chicago program, unlike similar programs at MIT, Carnegie Mellon and Caltech, focuses on the biological side of neuroscience rather than the robotics side. "It really makes our program unique, and it prepares undergraduates and graduate students for careers in the field by giving them a strong foundation," Ulinski said. Of the two models of computational neuroscience, one focuses on producing solutions for the engineering and computer science industries while the other focuses on how the brain works. The computational neuroscience group here is in the latter category.

The program begins with a seminar series featuring lecturers from a broad spectrum of departments. "There really has to be an interdisciplinary approach to this, because learning how the brain works involves biology, psychology, math, physics and more," Ulinski said. The lectures cover such topics as how the brain can be imaged and how neurons are studied in the laboratory. The program also includes a group of core courses, including an introductory class on neuroscience, a neuropsychology class that explores brain damage and a mathematics class on computational techniques used in connectionist modeling.

Although a degree in computational neuroscience is not offered by the University, many of the undergraduate students participating in the program have said it has given them a good foundation for entering the job market or for going on to graduate school. The program also attracts graduate students who believe that the integrative approach offered here will prepare them to specialize in a variety of fields later on, according to Ulinski.

"The next step," he said, "is to compete with other computational neuroscience programs by advertising and attracting more students."