A bidirectional associative memory explanation of posttraumatic stress disorder

Abstract

Network theory is a relatively recent explanatory development to improve our understanding of emotions and emotional disorders. Connectionist neural networks are introduced as mathematically well-defined network theories that have been suggested as PTSD models, partly because of their ability to form and retrieve memories. Explanatory requirements that a comprehensive PTSD theory must have are reviewed. A connectionist neural network system called the bidirectional associative memory (BAM) is extended to encode emotion and cognition and is then shown to satisfy all PTSD explanatory requirements and consequently constitutes a comprehensive PTSD theory. Empirical work pertaining to a fundamental assumption that learning alters brain structure in ways that can be detected by neuroimaging is reviewed. Novel predictions are made, and relevance to neuroscience is discussed.