Front Matter Free

The editor acknowledges the enormous administrative effort from Dr. Kara Rader in making this book a reality.

In Nordstrom (2021), we presented the background and building blocks associated with quantitative imaging (QI) in the medical arena. The cornerstone to accepted utility for QI is the application of standards to the entire imaging process. Everything, from the data collection process through to final image display, must be under the control of standard operating procedures (SOPs); Chap. 2, by Kinahan et al., and Chap. 3, by Guimaraes, of Nordstrom (2021) emphasized this necessity. Standards are an important part of multisite clinical trials, where detailed protocols are established to harmonize results across the collaborating sites. Although, for these cases, compliance to SOPs is relatively easy to enforce, this is much less the case in standard-of-care community practice. Reluctance to comply to SOPs is a drawback to rapid acceptance of QI methods across all medical practice.

The tools typically used in QI were then presented. The RECIST method was an early contributor to the movement from qualitative evaluation of medical images toward providing some level of quantitative results. This longest-dimension measurement of tumors was successful in predicting patient outcome for some tumors, but ran into difficulty with more diffuse or fibrous images of tumors. However, other methods for extracting quantitative information from medical images have become popular and productive in recent years. For example, radiomics and artificial intelligence (AI) are now at the forefront of image analysis. The potential of these methods was presented in Nordstrom (2021), and is enlarged upon in various chapters of this book.

However, the ultimate value of these tools rests on their ability to determine exactly where the disease is located within the image, and to define its margins. The process of image segmentation is so critical to QI that it was given its own chapter in Nordstrom (2021). The process of defining the disease location precisely, with accurate boundaries, is more difficult than we might think. The boundaries of many tumors and organs can be indistinct, and can share tissue properties with the background. Partial volume effects can further blur these boundaries. Yet, the importance of making accurate boundary locations is critical to accurate QI methods that extract information for processing.

Nordstrom (2021) ends with the recognition that QI is not a single application of a specific method to a single image or series of medical images. It is a process of methods, each with its own set of SOPs and input/output formats. The creation of pipelines is a way to link methods together in a purpose-driven package. The pipeline concept allows for flexible selection and configuration of processing steps, bringing QI closer to user acceptability. It also allows for rapid and consistent processing of imaging in bulk, increasing the throughput of image processing in clinical trials.

This book takes the principles and caveats of QI presented in Nordstrom (2021) and applies them in a number of ways. First, QI is presented as part of the various imaging modalities used in the clinic. Magnetic resonance imaging (MRI), positron emission tomography (PET), computed tomography (CT), and ultrasound (US) are clinical imaging methods where QI is making inroads. These imaging modalities are presented, with emphasis of the role of QI in each.

But emphasis on modalities is only one way to highlight the roles of QI in medicine. Looking at various organ sites to see how QI is applied, where the difficulties lie, and how they are being overcome, is yet another way of spotlighting the role of QI in medicine. Different chapters focus on the breast, brain, lung, bladder, and head and neck, to demonstrate the ways that QI is used in each.

Of course, it is not possible to separate completely the two approaches for discussing QI in medicine. A discussion of brain imaging will naturally bring up the various forms of MRI, and, conversely, a discussion of QI in MRI will focus, at least in part, on brain imaging. The goal in this book is to be as inclusive as possible of all aspects of QI in medicine, without worrying about the inevitable overlaps. As a result, the reader is encouraged to look at both the image modality chapters and the organ site chapters for a complete comprehension of QI in specific areas.

The lingering question regarding QI in medicine is, what is its clinical utility? Yet this question remains, despite all the definitions of QI and imaging biomarkers, all the presentations of algorithms and tools, and all the evidence that QI shows promise in a number of tumor sites. It is certainly one thing to demonstrate abilities in a research setting, but it is quite another to demonstrate clinical utility. There is no doubt that QI comes to the clinical environment carrying a lot of baggage, in the form of additional SOPs, quality assurance necessities, and reporting requirements. So, while there may be areas where the automation or semi-automation of image analysis can be attractive, the extra burden of standards and quality control can negate that attractiveness.

A constant issue from the radiologist community is concern over the validation of QI results. Without robust validation of algorithms, the weight of procedures and requirements associated with QI will be greater than the promise of more efficient clinical decision support. This is especially true when advancing AI or deep learning methods that are not supported by physiologic or metabolic models of disease. After all, in the words of Arthur C. Clarke, “Any sufficiently complex technology is indistinguishable from magic,” and radiologists are not going to begin relying on magic for patient diagnoses. This means that QI must prove itself through the process of validation in clinical trials.

To date, QI has established a firm foothold in clinical trial decision support. It will not go away. The question is, how rapidly will it be able to grow and expand in the clinical environment. This book, focusing on the applications of QI, is a strong indication that, although the acceptance of QI methods may be slow, there is steady progress in that direction.

Reinhard R. Beichel

Professor, Department of ECE, University of Iowa, Iowa City, IA

John M. Buatti

Department of Radiaton Oncoogy, University of Iowa, Iowa City, IA

Heang-Ping Chan

Paul L. Carson Collegiate Professor, Radiology, Director, CAD-Ai Research Laboratory, University of Michigan, Ann Arbor, MI

Jana Delfino

Center for Devices and Radiological Health, U.S. Food and Drug Administration, Silver Spring, MD

Laurent Dercle

Columbia University, Department of Radiology, New York, NY

Marios A. Gavrielides

Center for Devices and Radiological Health, U.S. Food and Drug Administration, Silver Spring, MD (currently with AstraZeneca, Gaithersburg, MD)

Mohammed Goryawala

Research Assistant Professor, Department of Radiology, University of Miami, Miami, FL

Lubomir Hadjiyski

Professor, Radiology, Rogel Cancer Center, University of Michigan, Ann Arbor, MI

Timothy J. Hall

Department of Medical Physics, University of Wisconsin-Madison, Madison, WI

Deep K. Hathi

University of California, San Francisco, 1600 Divisadero Street, San Francisco, CA, USA

Vaios Hatzoglou

Memorial Sloan Kettering Cancer Center, New York, NY

Nola M. Hylton

University of California, San Francisco, 1600 Divisadero Street, San Francisco, CA, USA

Ella F. Jones

University of California, San Francisco, 1600 Divisadero Street, San Francisco, CA, USA

Qin Li

Center for Devices and Radiological Health, U.S. Food and Drug Administration, Silver Spring, MD (currently with AstraZeneca, Waltham, MA)

Martin A. Lodge

Department of Radiology, Johns Hopkins University, Baltimore, MD

David Mankoff

Department of Radiology, University of Pennsylvania, Philadelphia, PA

Michael McNitt-Gray

Director, Physics and Biology in Medicine Graduate Program, Professor, Department of Radiological Sciences, David Geffen School of Medicine at UCLA, University of California Los Angeles, Los Angeles, CA

Robert J. Nordstrom

Director, Quantitative Imaging Network, National Cancer Institute, 9609 Medical Center Drive, Rockville, MD

Ramesh Paudyal

Memorial Sloan Kettering Cancer Center, New York, NY

Nicholas Petrick

Center for Devices and Radiological Health, U.S. Food and Drug Administration, Silver Spring, MD

Ivan M. Rosado-Mendez

Instituto de Física, Universidad Nacional Autónoma de México, Mexico City, Mexico

Ravi Samala

Adjunct Research Assistant Professor, Radiology, University of Michigan, Ann Arbor, MI

Kathleen M. Schmainda

Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI

Lawrence H. Schwartz

Department of Radiology, Columbia University, New York, NY

Akash Deelip Shah

Memorial Sloan Kettering Cancer Center, New York, NY

Hyunsuk Shim

Department of Radiation Oncology, Emory University School of Medicine, Atlanta, GA

Amita Shukla-Dave

Memorial Sloan Kettering Cancer Center, New York, NY

Richard L. Wahl

Chairman, Department of Radiology, Washington University, St. Louis, MO

Brent Weinberg

Department of Radiology and Imaging Sciences, Emory University School of Medicine, Atlanta, GA

Binsheng Zhao

Columbia University, Department of Radiology, New York, NY