Unsupervised Agent-based Artificial Intelligence Learning System

An Artificial Intelligence System that Mimics the Problem-solving Pattern of the Human Brain

Invention Summary

The Agent Based Brain Model, or ABBM, uses brain connectivity information generated from imaging modalities such as fMRI that enable the interference of functional interdependence between brain regions. Functional connectivity is used in establishing brain networks comprising the links between specific brain regions called nodes. Interactions between various nodes reflect human brain architecture. This innovation inputs human imaging data into the system and results in a computer-based artificial intelligence system that has the potential to produce emergent solutions in a fashion similar to the human brain.

Market Need

Current artificial intelligence systems are developed based on previously learned or programmed responses to specific stimuli that are different from the methods used by the human brain. These systems preclude the use of emergent behavior in problem solving. Researchers at Wake Forest Baptist Medical Center have developed a device with a form of strong artificial intelligence that has demonstrated the ability to solve computational problems and exhibit emergent behaviors. This computer-based system remembers the behaviors that it has produced before and the benefits of those behaviors while being able to interact within an environment.

Competitive Benefits

  • ABBM develops emergent behaviors dependent on environmental experience
  • ABBM uses emergent behavior in problem solving, a characteristic essential to mimicking human brain function
  • ABBM adapts to new environments independently of external control and performs unsupervised learning without external algorithms


  • In a clinical research program to model brain behavior in disease states, in interaction with potential drugs, in neurological conditions, in injury and in psychological therapy
  • Applications in personalized medicine for deducing expected prognosis and guiding treatment planning
  • A research tool to allow the modeling of how cognitive intervention such as “brain training/altered decision making/game playing” changes brain function


  • ABBM has successfully utilized input connectivity data generated from fMRI data and independent decision-making tasks
  • Working software for computation algorithms to run the ABBM model
  • Provisional patent application filed


  • Paul J. Laurienti, MD, PhD
  • Satoru Hayasaka, PhD
  • Karen Joyce

Related Publications

Joyce KE, et al. A genetic algorithm for controlling an agent-based model of the functional human brain. Proceedings of the Rocky Mountain Bioengineering Symposium, Mar. 2012

Moussa MN, et al. Changes in cognitive state alter human functional brain networks. Front Hum Neurosci.2011;5:83

Steen M, et al. Assessing the consistency of community structure in complex networks. Phys Rev E Stat Nonlin Soft Matter Phys. 2011 Jul;84:016111

Licensing Contact

Charlie Shaw, PhD
Licensing Analyst

Back to Top