Brains gather, store, and act on data gathered from their surroundings. This means that the data so taken in must be converted into information for further processing. On a physical level, we know that brains are composed of several types of specialized cells, generically called neurons, that carry out the gathering and processing of data captured by the senses. How the various elements of the nervous system gather raw data and process it is the subject of the study of neural information processing.

An introduction to neural information processing presents an overview of the current state of the field, drawing from a large number of sources. It traces the history of neuroscience, showing how parts of the brain were named, what early researchers thought those components did, and what we understand today of brain function. Experiments done and models developed over many decades of research are cited.

Beginning with the anatomy of the nervous system, the book proceeds to the components of the brain, ultimately focusing on the brain’s individual cells, from sensory receptors to neurons. A neuron is an electrochemical object that changes states through chemical reactions and electrical transmissions. In particular, neurons are arranged into a network through multiple connections, synapses, to other neurons. As a neuron gets inputs from those connected to it, it may either generate an impulse or be inhibited from so doing. Doing a statistical analysis and applying Shannon’s information theory to the impulses generated by a neuron over a period of time permits calculation of the amount of information it generates.

Once a measure of information exists and can be modeled, the next step is to model how the information is coded for transmission to the brain and other parts of the nervous system, for example, to cause motor actions to be carried out. With coding described, neural information processing and models of the various components in the visual and olfactory areas are discussed in detail.

The final chapter presents several neural network models, beginning with the classic Hopfield model of associative memory, and then proceeds to discuss continuous attractor networks, which model neural information representation; reservoir networks, which map input streams to output streams in real time; models for decision making; and finally models that incorporate dynamical synapses, that is, incorporating the real behavior of synapses, which modify their actions based on activity in neurons connected to their host neuron. In all models, mathematics are presented as needed to explain the model’s behavior.

The book is well organized, leading as it does from the physical system being observed to the lowest level, then up to circuits and networks, along the way describing various models that attempt to describe the behavior of the physical system. Readers should understand calculus for a deep understanding of the mathematical models. As with other Springer books, the authors of this one appear to have English as a second language. The writing is generally clear and readable, but there are some instances of unusual word choices. Curiously, there is no index. The book should be of interest to anyone interested in the current state of neuroscience. It covers quite a bit of ground in a concise but readable manner.

More reviews about this item: Amazon