Innovations in medical imaging

The very name ‘Johns Hopkins’ means medicine to many Americans. Much of what we take for granted in medicine today – from the rigorous training of physicians and nurses to the emphasis on research and the rapid application of that research to patient care – emerged from innovations made when it was a brand new medical center in Baltimore, more than 100 years ago.

The center now uses one name – Johns Hopkins Medicine – to identify its whole medical enterprise. This $4 billion organization comprises the physicians and scientists of The Johns Hopkins University School of Medicine and the health professionals and facilities that make up the broad Johns Hopkins Health System.

John Hopkins has not rested on its laurels: the organization is still at the forefront of medical innovation. As James Philbin, Senior Director of Medical Imaging, explains: “The thing that distinguishes us most is that we have a group of world-class researchers and clinicians who are dedicated to excellence. In the medical imaging department, several years ago we realized that IT was the foundation of modern radiology, and we now think of imaging informatics as the fundamental modality that all the other modalities depend on.”

No more film
One innovative program the department has undertaken is the Filmless 2006 Initiative. The goal of this initiative was to eliminate the use of film for radiology images. “Our aim was to be completely filmless and paperless by the end of the year 2006,” says Philbin.  “I’m proud to say we achieved that. The program resulted in about a 17 percent return on investment inside radiology, but one of the things we realized during the initiative was that it made sense to do an enterprise archive, and we immediately included some of cardiology and radiation oncology and the radiology department at Bayview Medical Center, a sister hospital. When we included those, the ROI went up to 45 percent.

“Then the financial people at Hopkins did an analysis on the soft gains, such as less patient time in bed, less time hunting for lost films, the fact that more than one physician could look at films at the same time. When we looked at that, the ROI went up to 75 percent. So the program has actually paid for itself in the two years since we did it.”

One of the biggest challenges of this initiative was transforming the orthopedics department into a filmless operation. Special large monitors had to be installed, so physicians could see spines and long limbs, and the same thing held true in the ORs. Philbin says they are all filmless today, though occasionally some physicians do want to have film printed for the OR. In general, though, most surgeries are done using the standard LCD monitors in the ORs.

This process also led to greater collaboration across the organization. “During the process of going filmless, we realized that there would be many more economies of scale if we built an enterprise image archive, rather than just a radiology archive, and so we established one in the beginning of 2006, after we discussed things with several different departments, who agreed to store their images in the same archive.

“At this point, the archive contains about 200 terabytes of medical images and their associate reports and all of the different medical image producers in the hospital have agreed to join the archive, with the exception of pathology, which is considering it but hasn’t yet committed.”

Among other innovations, the department has a radiology information system (RIS) and a speech recognition system. They also have several remote rendering solutions that allow the images to be rendered in the data center and then transmitted over low-bandwidth lines, so that 3D can be done over broadband connections, which has enhanced the ability to read imaging studies from home, and to do teleradiology.

Future directions
In terms of future plans, Philbin’s biggest hope is to build a Medical Image Storage Grid (MISG), which would handle medical image life cycle management.  The MISG would eliminate the need for migration from one PACS vendor to another. It would also allow access to images at a much lower cost with a much lower requirement for human management and it would allow different applications to share DICOM objects, eliminating the need for separate caches for remote rendering systems. The Medical Image Storage Grid would use low-cost content addressable storage, which makes it easier to manage the objects. Philbin expects to expand the MISG to include waveform, genomic and proteomic information.

“The other big initiative we have for the next 18 months is remote rendering based on commodity hardware; so within that time period, all of our images will be rendered remotely, whether it’s 2D, 3D, or 4D. Since different remote rendering systems offer different functionality, we expect to have several. We’re also hoping to have plug-in image viewers for our other medical systems, such as EPR – our homegrown EMR. We also have Eclipsys Sunrise Clinical Manager, and we use Microsoft’s Amalga, so we plan to have plug-ins for those three systems that would allow them transparent access to images and reports from the image archive.

“Making the move from film to digital images combined with having an enterprise archive that allows us to store images and reports from various departments in the same archive, has created a research opportunity that has never existed before and that I expect will yield significant new knowledge.” 

Hopkins has recently created the Center for Medical Imaging Informatics, an interdisciplinary research center that will be focused on using the information in the archive for large scale medical imaging research projects; for example, it will be possible to mine not only the data in the reports, but also the information in the images themselves, using techniques such as organ recognition and registration. It will also focus on correlating information in the images across disciplines such as radiology, cardiology, oncology and pathology, something that has never been done at this scale before.

James Philbin is Senior Director of Medical Imaging at Johns Hopkins Medicine, responsible for enterprise medical imaging in the health system and the medical school. He is also in charge of radiology IT at Johns Hopkins. He has worked in IT for over 20 years, spending ten years as a computer science researcher at the NEC Research Institute in Princeton, New Jersey, and has been the CEO of two high tech startups.