Introduction

During the past several decades, many diagnostic cardiovascular modalities have been developed. Combining these into multimodal imaging (MMI) creates enhanced decision support for clinical management of patients with cardiovascular disease.

Progression from additive consideration of the primary data from different modalities, to side-by-side comparison of the different images, to superimposition of the structural and functional characteristics of the different images, provides the increased understanding required for optimal improvement of clinical diagnosis. Multimodal imaging is also a way to continually update and reevaluate paradigms of individual diagnostic methods.

In this book, a team of distinguished authors—clinicians as well as computer scientists, mathematicians, and physicists—has been brought together to provide a wide range of aspects important to the development and use of MMI. The authors have been requested to confine their chapters to their own area of interest and expertise and to include both their current experiences and future considerations.

The chapters are divided into three sections based on the perspectives of the authors. Sections I and II contain chapters that focus on a single diagnostic modality and its broad application in clinical cardiovascular conditions. Section III contains chapters that focus on a single clinical condition and how MMI can provide decision support. This format provides duplication of information in chapters in the sections of the book.

In Table I–1, the imaging modalities included in the 17 chapters in Sections I and II are presented as columns, and the general clinical conditions for which they provide diagnostic capability are presented as rows. The modalities are ordered according to the sequence of chapters and are numbered from 1 to 17 accordingly. Some of the modalities are considered in more than one chapter in Section I.

Chapters in Section I include many modalities that currently produce direct images of cardiovascular structures and their function (eg, coronary angiography, echocardiography, SPECT, and magnetic resonance imaging). Because the inexpensive and widely available electrocardiogram (ECG) is now being developed into an imaging modality, several chapters are devoted to considering the perspectives of standard scalar ECG, spatial dipolar electrocardiotopographic [DECARTO] imaging, the inverse problem of ECG (ie, deducing cardiac excitation patterns from body surface recordings), and autonomic modulation. Section II covers the emerging technologies that are leading to the creation of direct imaging of cardiac activation and recovery. This requires extensive mathematical modeling by physicists, engineers, and computer scientists in concert with the academic clinician-investigators who will ultimately use the methods for their diagnostic and therapeutic decision support. Three of the imaging modalities that exist as simulations or models, rather than as clinically available methods, are considered in dedicated chapters: "Cardiac Simulation for Education: Electromechanical Modeling Applied to Cardiac Resynchronization Therapy" (Chapter 13 by Constantino et al), "Computational Cardiac Electrophysiology: Modeling Tissue and Organ" (Chapter 14 by Bishop et al), and "The Electrocardiogram According to ECGSIM" (Chapter 15 by van Oosterom et al). These methods are currently used to guide clinical investigators in the development and evaluation of new cardiovascular therapies and might be developed into clinically useful tools in the future.

Some explanations about the various modalities might be helpful. Coronary angiography refers only to radiographic visualization of intra-arterial contrast. The chapters on the common general clinical imaging methods (eg, echocardiography) include the many specific adaptations in which those methods are employed (eg, imaging of blood flow based on the Doppler principle; conventional precordial, transesophageal, and intracardiac imaging).

In Table I–2, the eight chapters presenting the use of MMI for decision support in the various clinical conditions in Section III are presented in the seven columns, and the individual modalities are presented as rows. The clinical conditions are ordered according to the sequence of chapters, which are numbered from 18 to 25. Some of the clinical conditions listed in Table I–1 do not appear in Table I–2 because in Section III, no chapter dedicated to that condition (eg, valve disease) has been included. Also, some of the imaging modalities listed in Table I–1 do not appear in Table I–2 because none of the authors in Section III has elected to consider that modality (eg, optical imaging).

Chapters in Section III emphasize how MMI can be used for diagnosis and therapeutic decision support for various cardiac conditions (eg, congenital heart diseases, ventricular cardiomyopathies, myocardial ischemia/infarction). A broad spectrum of cardiovascular conditions has been included with no attempt to fit the size of the chapters to the clinical importance of the conditions. Rather, the authors selected the breadth to include in their chapter based on both interest and expertise.

Some chapters provide case reports of how academic clinicians/scientists are currently using MMI modalities to improve their clinical therapeutic decision support (eg, Chapter 21 by Ubachs et al).

It should be emphasized that the pathway toward successful integration of information from multiple imaging modalities is challenging. A common question is how to consider discordant results when performing combination imaging. The MMI approach extends beyond the traditional concept where one method is compared with a gold standard method to the new concept that each modality provides unique diagnostic information. Traditionally, when discrepancies with the gold standard modality occur, the term "false" is often used. The discrepant results are assigned as false negative or false positive, and these false results may diminish the perceived value of a certain modality in the diagnostic decision making. The alternative consideration in MMI is that different diagnostic imaging modalities are based on different biophysical principles. Therefore, they each provide different information and visualize different specific characteristics of the same structures or functions. Thus, both agreements and disagreements between the results of particular modalities may have their own specific diagnostic value.

The perspective of a step-wise approach to the means of combining several imaging modalities to provide understanding of cardiovascular pathophysiologic processes is presented in Figure I–1 and described in the following sections.

Adding Information from Multiple Imaging Modalities

Each modality provides its own means of displaying primary information about the heart, and each method is interpreted separately. Selected aspects of the result of applying the various individual modalities are then sequentially related. For example, to the information about cardiac dysfunction by echocardiography, a myocardial scintigram may add the information that the dysfunction is caused by reduced myocardial perfusion.

Side-by-Side Comparison of the Information from Different Imaging Modalities

This approach provides a tool to test the accuracy of a more available method comparing it to the less available method considered as a reference. As pointed out earlier, it needs to be kept in mind that different diagnostic imaging methods are often based on different biophysical principles and thus provide more or less different information and visualize different specific characteristics of the same structures or functions. Disagreements between the results of particular methods may therefore have their own specific diagnostic value.

Conversion of a One- or Two-Dimensional Imaging Modality into Three-Dimensional Myocardial Anatomy

Using sophisticated methods of image processing, the recordings of one imaging modality can be transformed into three-dimensional (3D) images. This conversion can provide better understanding of the structure and function of the heart in an individual patient. An example is the conversion of the two-dimensional echocardiographic images of the heart valves into 3D images.

Converting an ECG into an Image for Side-by-Side Comparison with Other Imaging Modalities

An ECG can be projected onto a 3D image of a digital heart model and then compared with information from other imaging methods. This provides a strong tool for understanding electrofunctional and anatomic characteristics of the heart. An example is the projection of an ECG showing infarction into a DECARTO plot that indicates size and localization of the infarcted area. The DECARTO plot can then be directly compared with a method for assessment of myocardial perfusion (eg, SPECT imaging).

Fusion of Information from Multiple Imaging Modalities

Fusion of images may make the resulting image more informative than any of the input images. An example is the superimposition of the 3D model of the infarcted area by one modality onto an initially ischemic area by another modality for assessment of myocardial salvage.

Rapid development of new imaging modalities is opening extensive opportunities in imaging the cardiovascular system from anatomic, pathophysiologic, and electrophysiologic perspectives. McGraw-Hill publishes the cardiovascular text Hurst's The Heart, and this book on cardiovascular MMI can be considered as a complementary volume. The mission of this book is to introduce several levels of integration of multimodality information—with the patient both at the beginning and at the end of the MMI chain.