The Handbook of the Medical Image Perception and Techniques. Ehsan Samei and Elizabeth Krupinski (Editors). Cambridge 2010, 436 pages, 170 illustrations, $169.00.

For those interested in delving into issues related to image perception, this Handbook serves a useful purpose because it address topics which radiologists seldom consider. There are six sections of this book: Historical reflections and theoretical foundations; Science of image perception; perception metrics; Decision support and CADs; Optimization of image systems; Future of medical image perception. Twenty authors have contributed to these chapters.

With the increased importance of imaging informatics (IT) and the establishing of sections of IT in many departments, the information and concepts contained in this book are timely. The book begins with introductory chapters defining the basics of medical imaging perception and the history of the development of concepts in perception. One quickly becomes acquainted (or re-acquainted) with the idea that radiologist errors and intraobserver variability is significant. From that baseline the book builds a case for understanding the multiple factors which influence what we can see and what we think we see. Reported observational variability between radiologists (around 30%) is not unique to medical imaging and parallels to other areas of science can be found. Subjecting all of this mathematical (statistical) analysis and rigorous investigation has lead to an important component of imaging informatics/imaging evaluation. The historical background of many items, such as the logical development of ROC graphs, provides interesting reading.  Studies on visual search (how the eye tracks an image) and gaze patterns relative to medical experience shows the enormous variations in the evaluation of medical images.

While some material will hold minimal or no interest for the members of the ASNR (such as the signal detection chapters or surrogate to perception experiments) other areas will (such as the role of expertise in image interpretation, or estimation vs. measurement-[a hot topic since there is push for quantitative reporting], CAD information  both from the perception standpoint and in the design of CAD studies, CADs in chest and mammography, CAD accuracy, CAD risks and benefits and perceptual issues in mammography. The chapter on optimization of 2D and 3D imaging systems is too intense except for the diehard image scientist. Ergonomics in the radiology reading room, while short, is a chapter you can wave in front of your hospital or practice manager.

Those involved in imaging informatics will find this book useful.