The analysis of medical images is built on the foundational techniques of general image processing, but incorporates special techniques and methods to account for the nature of medical image acquisition, content, and requirements. This book is designed for professionals in the medical field familiar with anatomy and physiology, but requires an introduction to the analysis of medical imagery. As the title states, this book is a “guide,” written to serve not only as an introduction, but also as a reference and a jumping-off point to more detailed works on specific topics.
Imaging of the body is challenging; many unique imaging methods have been developed to produce useful images that show specific conditions and features. After a chapter introducing the general topic of medical image analysis and presenting several software tools, Toennies presents the most common medical imaging techniques: X-ray, magnetic resonance imaging (MRI), ultrasound, and nuclear imaging. The presentation is clear and concise; in particular, MRI is discussed at a good level of detail and quite clearly for such a complex topic. Other imaging modalities are briefly--perhaps too briefly--described; for example, optical coherence tomography (OCT) and other ocular imaging techniques deserve more than one page.
The majority of the book progresses through the usual steps of processing an image: storage and transfer, enhancement, feature detection, segmentation, registration, shape and image understanding, and classification. While general image processing principles are included, there is a strong emphasis on techniques useful for medical images. A chapter on validation is well written and especially important for the analysis of medical images. For the stated audience, the discussion of each topic is approachable, technically correct, and leads into the next section. I believe that medical professionals will learn much from these chapters.
Segmentation in particular is properly noted as crucial for medical image analysis and is well covered. One chapter introduces the basic problem and presents classical methods such as Otsu’s method and the watershed transform. Segmentation in feature space is described separately. An especially useful chapter on segmentation as a graph problem gives an introduction to the methods and a good list of references.
The book is well written and accurate. The author states that he has made a number of additions and corrections in this new edition; the result is very good. While the book is titled Guide to medical image analysis, it’s well suited as a textbook for medical professionals. I am evaluating it for adoption in a medical imaging course, and would recommend it to those in the medical field who want a detailed discussion of medical image analysis.