The human brain is a complex organ with functions and processes that remain a mystery for modern day science. Neurological dysfunction is often used as a measure (and is sometimes a cause) of a magnitude of disease, and deciphering the brain’s complex chemistry and function could enable a leap in our understanding of brain physiology and the approaches to treatment.
Randy McIntosh, a psychology professor at U of T, is leading an international project to create the world’s first operating virtual brain.
McIntosh is a senior scientist at the Rotman Research Institute, which has teamed up with scientists from Australia, France, Holland, Spain, and the United States to create a virtual brain that will copy patterns of brain activity when people are involved in a thinking task.
This project is made from software running on a Linux cluster and is similar to the genome project in several ways. “First, it catalogues the important dynamics of the human brain for a wide range of cognitive functions, and second, it provides a new perspective to understand why brain damage and disease have their particular effects,” explains McIntosh.
The virtual brain will catalogue dynamics of brain activity when performing mental tasks. The catalogue comprises a large sample of healthy brains from people of a wide range of age groups. Comparing patients’ brain activity with the corresponding healthy sample could help determine whether or not (and where) a person has a neurological problem.
The new machine utilizes cutting-edge technology and is a predictive modelling tool that will enable doctors to evaluate many brain disorders, such as cognitive impairment caused by Alzheimer’s disease.
The machine will work by allowing clinicians to effectively experiment with how a person’s brain would react under certain conditions. A patient’s unique neural structure is loaded into the virtual brain to find out how the model responds to the disruption of normal patterns and how it tries to re-stabilize. Looking at this reaction, a clinician will then try to identify the best options for treating the affected areas of the brain.
According to McIntosh, the process of creating this complex model has so far been slow but rewarding. “The collaboration has given us new insight into the relation between brain structure and function that we did not know before,” he said.
The project brought together a multitude of resources and data, including functional MRIS, structural MRIs, and demographic information. The team then archived the data in an accessible database, and integrated it with the computational model.
One of the most formidable challenges for McIntosh will come when the team starts creating virtual models that capture mental functions, such as seeing, hearing, and remembering.
“We will be able to identify the structural foundation and have the virtual brain produce the exact same pattern for hearing a piece of music that you would see in a human. The question is, does the virtual brain experience that music the same way the human does? This gets into a philosophical realm that we always face in this business.”
The different parts of the virtual brain are located at most of the collaborating sites, while the database is stored in the RRI. The virtual brain is currently only accessible by scientists, but will be available to a broader audience if it reaches the stage when it is comparable to a real human brain.
A similar project called the Swiss Blue Brain models cortical columns (structured columns composed of thousands of neurons), while the virtual brain models the interaction between several cortical columns. “It is a race of sorts, but we are not heading for the same finish line,” explains McIntosh.
