Although receiver-operating characteristic (ROC) methodology has become ubiquitous in the evaluation of medical imaging technology, this was not always so; in fact, the fundamental insight that detection or classification performance can be described most meaningfully in terms of a variable trade-off between true-positive and false-positive rates did not emerge in any field until around 1950. John Swets was there at the beginning, initially as a graduate student in psychology at the University of Michigan and subsequently during an illustrious career that included substantial research and managerial accomplishments in academic, corporate, and public settings. As its subtitle indicates, this memoir traces the parallel paths of Dr Swets’s career and development of signal detection theory (SDT), which provides the conceptual basis on which ROC analysis has been built.
The “tulips” of the book’s title refers to Dr Swets’s Dutch heritage and early life in Holland, Michigan, whereas the plural “thresholds” harks to his (and our) conceptual transition from one kind of “threshold” to another. Until the mid-20th century, human perception was thought to involve a fixed threshold such that no sensory input below some minimum level of stimulation produced a corresponding sensation; thus, false-positive detections were ascribed to malfeasance or hallucination. SDT posited, instead, that sensory “signals” must be detected within a context of underlying “noise”—inherent statistical variation—that could both mask real signals and mimic signals when none were present. From this perspective, a signal is “detected” or a difference between two kinds of signals is “recognized” by comparing the statistically variable stimulus with a variable decision threshold that the human observer (or a machine observer) can, in principle, adjust to accommodate different levels of signal prevalence and/or different costs of false-positive and false-negative errors, and so on. To make a decadelong story short: SDT won out and now provides both the conceptual and the quantitative basis for formally making and evaluating most decisions that involve statistical uncertainty.
Although radiologists and medical physicists who do not know Dr Swets personally are likely to begin reading this book mainly for its gracefully told story of SDT and ROC analysis, most will find themselves enjoying the author’s nostalgic yet penetrating memories of growing up in a simpler time, his gentle humor, and his remarkable recall. Most readers will also find themselves engrossed by Dr Swets’s firsthand accounts of responsibilities and personal interactions that ranged from basic research in psychophysics to the birth of the Internet, from the idea of estimating ROC curves by collecting confidence ratings (without which ROC analysis would be merely a concept in medical imaging) to serving as senior vice president of Bolt, Beranek and Newman, Inc, and from the creation of a standardized language for reporting interpretations of medical images (with Carl D’Orsi and Daniel Kopans, thereby anticipating the Breast Imaging Reporting and Data System) to service on the National Research Council’s Board of Governors.
To whom is this book most likely to appeal? In attempting to address that question, I cannot hope to better the last paragraph of Dr Swets’s foreword: “Probably the best audience for this book consists of those investigators who have come to use the ROC as a disembodied statistical technique. They may now find it attractive to learn of the history and scientific grounding of the concepts that support the ROC through an insider’s memoir.”
Book:
Contents: ★★★★
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Grading Key
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<