Veterans reflect on 25 years of Diagnostic Imaging

November 5, 2004

The first edition of Diagnostic Imaging was published in San Francisco 25 years ago this month. The 24-page tabloid publication was created as a "news journal of radiology, nuclear medicine, and ultrasound." Many changes have been made since then-in content, focus, layout, and design-but the news format has remained.

The first edition of Diagnostic Imaging was published in San Francisco 25 years ago this month. The 24-page tabloid publication was created as a "news journal of radiology, nuclear medicine, and ultrasound." Many changes have been made since then-in content, focus, layout, and design-but the news format has remained.

That format has proved to be fortuitous. News, after all, is the recording of the process of change, and radiology has consistently been one of the most dynamic medical specialties. Diagnostic Imaging asked four of its editorial board members to contribute brief reflections on the practice of radiology in 1979, and how it has changed since then. Following are their responses.

Past 25 years have been great for patient care

Dr. James Thrall, radiologist-in-chief, Massachusetts General Hospital

Maybe the best way to frame the changes of the last 25 years is to look at what is different for our patients, referring physicians, and other stakeholders. One of the hospital vice presidents at MGH was complaining recently about the growth of imaging exams. She pointed to the higher numbers of ultrasound, CT, and MRI procedures and asked, "Didn't anything go down?" Our response from radiology was, "Yes ma'am, exploratory surgery!"

The appendicitis story is a compelling example. After 100 years of never achieving better than 75% diagnostic accuracy by physical exam and conventional workup for patients going to the operating room, adding the right imaging has increased that to 97% to 98% in good hands. Just think of the impact on our patients. What a wonderful outcome from advances in imaging. Pick any organ system, and there is a similar example.

When radiology can't eliminate the need for surgery, we can often minimize its more negative effects by using image-guided methods. It is interesting that both the surgeon and the interventional radiologist need to visualize disease in order to treat it, but conventional surgery does this by destroying normal tissue, thereby increasing morbidity and risk of mortality without adding therapeutic value.

Minimally invasive methods represent a sea change in medical practice, to the great benefit of patients, but they have also brought what surgeons do and what radiologists do closer together, thus igniting turf wars that are destined to go on for many years. When DI was getting under way 25 years ago, our big turf battle was about obstetrical ultrasound, and we were ripped that our obstetrical colleagues thought they could do it. Today, everyone and their little brother does ultrasound, but radiologists still perform more than anyone else. Maybe that's what will happen with image-guided therapies.

As a first-year radiology resident, I realized that the film library was a vital but mysterious, annoying, frustrating thing. Years later, when I became a department chair, I realized that I had greatly underestimated just how mysterious, annoying, and frustrating film handling could be. When we turned on our radiology information system at MGH in 1988, we realized that we were losing about 10% of the cases before we had even dictated them. Where did they go? I think many of them are still behind the lockers of the surgical house staff, and I am waiting for one of the hospital's walls to fall off under the burden.

So, DI, on your watch and mine, this problem evaporated into a sea of electrons. The film management problem that was never really solved is simply no longer an issue. Images are everywhere all the time, and productivity has skyrocketed. Interestingly, if one reads carefully the recommendations of the Institute of Medicine and reviews the thinking of quality gurus, it becomes apparent that radiology is a leader, even an exemplar, for the thesis that infrastructure and systems are the secret to sustainable quality improvement in healthcare, because we were there first with RIS and PACS. The rest of the medical world is trying to play catch-up with electronic medical records.

In 25 years, the abstraction of radiology has not changed. We are still engaged in exposing biological tissues to different sources of energy and recording the results of the interactions. On the other hand, practically every aspect of how we do this has changed dramatically. We have new energy sources. Our measuring stick that stopped at millimeters extends to molecular dimensions. Our coordinate system has expanded from two dimensions to three dimensions and more. Analog has given way to digital. But the most dramatic change on your watch, DI, has been the shift of radiology from an ancillary service to a position astride the critical path of care for most acute diseases and conditions. I see a future where patients will literally enter hospitals through round and tubular structures. Well, maybe in another 25 years.

November 1979: Will NMR tomography replace CT?

Dr. William Bradley, radiology chair, University of California, San Diego

In November 1979, I was in the middle of my second year of radiology residency at the University of California, San Francisco. Six months earlier, on Tuesday, May 15, Larry Crooks, a UCSF physicist, had made a presentation to a small group of faculty discussing a new technique called "NMR tomography." Although I was not at the meeting, my friend and mentor Steve Ross gave me the journal articles handed out at the presentation, all of them physics papers with a lot of triple integrals. I vaguely remember him challenging me to explain what was in those papers, which was somewhat easier back then, only six years after my Ph.D., than now after more than 30 years.

As I read the papers that evening, it was like being hit by the proverbial thunderbolt. I pulled out my battered copy of Roberts and Casserio's Organic Chemistry, which I had used as an undergrad at the California Institute of Technology in an attempt to understand how nuclear magnetic resonance (NMR) could be used to make an image. As I began to understand NMR tomography, my excitement grew. I stayed up until 4 a.m., which I don't think I had ever done before (or since) for anything . . . academic.

I was rotating through pediatric radiology. Charles Gooding and Bob Brasch required a three-page research project as part of the rotation. I decided to summarize how NMR tomography worked (yes, in three pages). A few months later, I realized that my initial explanation was so lame that I removed it from the official notebook of research projects.

Over the next year, I made frequent trips down to the UCSF Radiology Imaging Lab in Oyster Point, South San Francisco, to sit at Larry Crooks' knee and ask questions about NMR tomography. After each trip, I'd write a summary of the conversation so I wouldn't forget it. On each subsequent trip, we'd check to see if what I had written was correct.

In November 1979, I was rotating through the San Francisco VA Medical Center. I was asked to give what would be my first talk on NMR to the other residents. It was at noon. They all fell asleep. I realized I had to work on my material and technique. More trips to Oyster Point, more detail to fill in the gaps.

Personally, the timing of the clinical birth of what would eventually become known as MRI could not have happened at a better time. I had already read Paul and Juhl, so I could take resident call, and I was still a year away from my last year and the neurosis of oral boards. So I had a year to try to figure out how MRI worked.

At the end of the year, I was asked to give a talk at a national technologist meeting sponsored by UCSF. I agreed if they would pay for an artist to do drawings for a monograph, NMR tomography. That mimeographed monograph became the basis of my first chapters on the physics of MRI.

During the next 25 years, we bought the first commercial MR scanner based on the UCSF design (Diasonics, subsequently bought out by Toshiba) at Huntington Medical Research Institutes in Pasadena, Larry Crooks retired as a gazillionaire from his patents on MRI, and I followed Anne Roberts (daughter of Jack Roberts of Roberts and Casserio) as chair of radiology at UCSD, where we have just ordered eight 3T magnets. Life is good.

A traveling veteran looks back at 1979

Dr. Edward Staab, Wake Forest University

The editor of Diagnostic Imaging invited me to recollect some of the most memorable aspects of radiology in 1979, when the journal first started. He apologized for not getting to me sooner but did not have my most recent contact information on file. It is true that I have moved around somewhat in my academic career. Following a two-year sojourn defending our eastern shores against imminent attack while in the U.S. Army, I have spent time at four different medical schools, all in the southeast part of the country, and a few years in various capacities with the federal government at institutes within the National Institutes of Health.

The DI editor wondered if I was enrolled in the witness protection program, but I assured him that these moves were all for positive reasons. I mention this just so the unwary reader will know where my recollections stem from.

- CT was the new method on the block. We were justifying its existence as the technology improved from month to month. GE had recently introduced its 7800 machine with its revolutionary xenon detector. We argued about whether the translate-rotate method or the GE method was better. We occasionally used intravenous contrast with the studies. I made a trip to Thailand to introduce radiologists in the Far East to CT technology. Bangkok had the first CT scanner in Asia.

- MRI was in its infancy. The magnets came in different varieties ranging in size from 0.15T all the way up to 0.5T. We spent much of our time learning how to operate these devices without hurting the magnets or ourselves. Real excitement stemmed from a quench of a magnet. The noise was terrifying. Like CT, the earliest applications were in brain imaging.

- Nuclear medicine was relatively sophisticated and the main alternative technology to radiography in most departments. Digital cameras were still in the research environs of the manufacturers. We were still using Polaroid film, and liver-spleen scans were one of the most common studies. PET imaging was talked about at all the major nuclear medicine meetings but was thought to be a research tool and completely impractical for patient care.

- Mammography was improving but still being done with variable quality across the country. It actually was performed in many nonradiologists' offices.

- PACS was a figment of a few people's imagination. I worked in the computer science department at the University of North Carolina one summer, and I remember helping to create one of the very first 3D images of a lung tumor with the "beams-eye" view of the radiation therapy machine. I had to outline every image by hand. We then asked a computer that occupied a whole floor to calculate the information so that we could display it. It took two months of machine time.

- Entrepreneurial radiology probably existed, but I sure was not aware of it. It was anathema to advertise our services other than through personal contact and at educational or scientific meetings.

- Medical students were not very attuned to radiology. In fact, radiologists were seldom seen as teachers in the first few years of medical school except in a few medical centers. Radiology as a profession was not held up in high esteem to the medical students.

- Our residents spent far too much time worrying about their board examinations. Some things never change.

- The RSNA, American Roentgen Ray Society, and Society of Nuclear Medicine meetings were well attended and the mainstay of our educational process. A few CME courses were put on by major institutions, but the myriad CME courses offered today did not exist.

- Most departments were divided into radiology, radiation therapy, and nuclear medicine sections. Nuclear medicine and radiation therapy were separate because most chairs did not know much about those disciplines. The leading section was radiology.

sigma There was much discussion about how departments of radiology should be organized: along disease and organ system lines or by modality. Much thought and discussion about this took place within departments and at national forums. This was mostly an issue for academic departments. The radiologists in practice just got the job done.

Early perspective on the evolution of interventional radiology

Dr. G. David Dixon, interventional radiologist, St. Luke's Hospital

By 1979, when Diagnostic Imaging asked me to join its editorial board, a class of special procedures had been anointed "interventional radiology" by Dr. Alexander Margulis. By then we were well into draining kidneys and bile ducts. Lung biopsies were common. I well remember being booed off the podium in 1975 when I passed on the news that Dr. Hans Lunderquist was doing percutaneous needle biopsies of the pancreas. It was great to have had the opportunity to be the first one to do so many new things. Early on, we often lived by the adage "Just because we don't know how to do it doesn't mean we shouldn't try."

During the early days, simple things had a way of naturally progressing. Antegrade pyelography became percutaneous nephrostomy when Dr. Seldinger's technique was applied. When the Grundzig balloon came along, we started dilating ureteral stenoses. A high point in my career occurred when I reported what was probably the first case of successful balloon dilatation of a ureteral stenosis.

The bread and butter of IR, diagnostic angiography, is disappearing fast. I learned my skills for intervention by doing these cases. Vascular intervention is so much easier using skilled hands honed by hundreds of those bread-and-butter cases. Fortunately, a new toy that provides real-time simulation of catheter manipulation is here to help young interventionalists hone their skills. I'm sure someone will soon be saying, "Can you imagine learning these skills on a real patient?"

So much has changed. Plain film is now computed radiography. Ultrasound now has 4D. Nuclear has PET. I've survived 40 years doing what I do and still love it, even if much of my work involves lowly venous access cases. It's been a great ride, and I'd do it all over again.