This nicely organized and well written text not only describes how to integrate modern computer technology into neuroanatomy; it provides a broadly applicable framework for that class of problems. Essentially a one-author text, it is consistent in editorial style and organization, so the reader hardly notices differences in the few chapters contributed by others. Illustrations are plentiful and I appreciated the occasional use of humor and cartoons to emphasize a point--especially when the concepts are difficult or excessively ponderous. Intended as a course text, and no doubt derived from course notes, it could be a useful guide or supplement to image analysis courses in neuroanatomy or computer science. The book is divided into seven major segments, which are unique and distinct from each other; a common bibliography integrates the volume, which is a helpful device in such a specific text.

The history of quantitative neuroanatomy is nicely presented in chapter 1. Even the casual reader will appreciate why neuroanatomists have turned to computer technology to help them in their research efforts. The historical treatment is necessarily brief, covering highlights without going into any area in great depth. The major figures in the field are included. Chapter 2 is a very broadly painted elementary description of computer hardware, while chapter 3 describes generic software concepts in a similar fashion. They provide a vocabulary, but not enough information to solve problems. The book presents what is available and is not a critical evaluation of the limitations of any particular piece of hardware for a specific problem. The author uses a very specific system for reference throughout the text, but its origin and rationale are not obvious.

Chapters 4, 5, and 6 are highly detailed descriptions of three data entry modes for defining neuron trees--from the microscope, by means of serial sections, and through the use of video technique. These chapters provide sufficient background and balance to allow exportation of the concepts to other areas; thus, scientists from other disciplines can begin to use the techniques with some appreciation of their limitations for a given application. Following surfaces and reconstructing in three dimensions are parts of modern scientific biology as well as clinical medicine. Chapter 7, on errors and their correction, and chapter 8, on methods of three-dimensional displays, are well written and present principles that will help a project succeed.

Chapters 9, 10, and 11 are nearly encyclopedic in the breadth of their coverage of mathematical techniques for describing and analyzing neural networks, including descriptive mathematics, topological analysis, and statistical analysis. While avoiding the pitfalls of providing too much specific mathematical or statistical detail, they are effective overviews. Readers who need a particular technique will find the references sufficient. These chapters are well written and, considering the complexity of the topic, easy to read and understand. Chapter 12 attempts to demonstrate how to build applications at this level. The presentation is too simple and assumes the reader has no or little knowledge of human-machine interfaces. That may have been true many years ago, but most modern systems have been designed to meet appropriate interface criteria. The key is to use the interface most appropriate for a specific technique and not attempt to either force a problem to a given solution or to impose an existing solution on a new problem.

Chapter 13 describes hardware and software techniques for image enhancement. It is too specific for general application; the techniques may be quickly outdated. Chapter 14 details the analysis of immunohisotchemical detail and gives factors influencing the accuracy and limitations of these techniques. Chapter 15 describes techniques for tracing neurons and even neural trees; this overview barely begins to hint at the hardware needed for this sophisticated application. These chapters present a broad spectrum of problems and solutions and offer a view of the future, but this view is not completely tempered by a sense of economic reality. Using hardware and software solutions to the problems posed has advantages, but one must question whether the increase in the processing volume is worth the incremental cost.

The last chapter is a catalog of “commercially available computer systems for neuroanatomy.” Not all of the systems are neuroanatomy-specific, and many of the applications go beyond anatomy in the strict sense of that term. Nonetheless, the chapter is a current list of image input and processing devices that can be used for any application.

This volume is reasonably complete, and most of it is sufficiently critical to be used as a text; some chapters will always be solid references, while others will require frequent updating as image analysis evolves and newer hardware becomes available. A chapter on clinical neuroanatomy and disease representation would round the volume out nicely.