| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
President's Address |
1 From the Department of Radiology, University of Pennsylvania Health System, 3400 Spruce Street, One Silverstein, Philadelphia, PA 19104. Presented at the 2002 RSNA annual meeting. Received March 24, 2003; revision requested April 28; revision received June 20; accepted June 25. Address correspondence to the author (e-mail: bryan@rad.upenn.edu.)
Index terms: Computers, diagnostic aid • Images, digitization • Picture archiving and communication system (PACS) • Radiological Society of North America • Radiology and radiologists
It is my pleasure and duty to deliver this address that traditionally updates or highlights key activities of the Radiological Society of North America (RSNA). The field of radiology or medical imaging is in the middle of a digital or computer revolution that is affecting all of science and its various fields, including medicine. I am very pleased to report that the RSNA and, in particular, one of its major committees, the Electronic Communications Committee (ECC), is playing a key role in this millennial change in medical imaging.
THE ECC
The ECC was founded, at least partially, at the instigation of Edward V. Staab, MD, in 1987 and was charged with the responsibility of maintaining radiology and the RSNA at the forefront of the rapidly evolving electronic communications technologies. This committee has been chaired by Dr Staab, Laurens V. Ackerman, MD, PhD, Carl C. Jaffe, MD, and R. Gilbert Jost, MD, and is currently chaired by Ronald L. Arenson, MD. More than 49 members of the RSNA have served on this committee, which has been one of the most productive and creative committees of our organization, having stimulated industry to implement the DICOM (digital imaging and communications in medicine) standard, created the infoRAD component of the annual meeting, and, most recently, offered the Integrating the Healthcare Enterprise (IHE) initiative. The IHE initiative has as its mission to stimulate the integration of health care information resources. The advantages of such an integrated enterprise are many, but can probably be summarized by the primary goal of the IHE program: To ensure that all the information required to make the best medical decisions be available to patients and their physicians when and where they need it. Clearly, in this age of high information density and the necessity of its efficient and rapid dissemination, the IHE initiative is a key program of the RSNA.
At the heart of IHE is the digital revolution we are now living. Two separate but related elements form the basis of this revolution. First are the new imaging modalities entering our practices that offer us more types of information or signals from the human body, and not just the traditional morphologic information, but physiologic and molecular data. These signals are increasingly detected in not one or two but three spatial dimensions and with increasing temporal resolution. Basically, we are able to make extraordinary, sophisticated, three-dimensional (3D) measurements of intrinsic biomedical importance in temporally dynamic fashions.
ANALOG-TO-DIGITAL CONVERSION
I will not focus further on these new imaging modalities, but will rather focus on the second basis of the digital revolution—digital data and technology. Essentially, all contemporary imaging studies consist of data that are intrinsically digital, allowing much more efficient and flexible use of these data and making them more readily available to our patients, referring physicians, and other partners in health care. The digital nature of these images will also allow us to evaluate them in a more scientific, quantitative fashion.
There are other rebellious characters driving this revolution, including (a) clinical demands, such as that for 24–7 coverage; (b) tight medical economics that demand more efficiency; (c) fundamental changes in the relationship between health care givers and patients who are becoming increasingly educated, enfranchised, and involved in their own health care; and (d) digital technology, particularly the World Wide Web, which provides the practical base for this revolution.
Easily obtained images of my brain’s anatomy and physiologic function during a simple finger-tapping task well illustrate the extraordinary power of digital imaging that will become increasingly apparent as we take advantage of the 
science of the bits.
New, refined signal and spatial analyses will yield richer biomedical data, and greater powers of visualization will derive from more complex computer-generated images. Seeing is believing is not only the motto but also the psychologic underpinning of radiology, and computer-assisted imaging will be more efficient, intuitive, and believable.
From my perspective, the first clinical digital images were the early CAT, or computerized axial tomography, scan images of the brain that were introduced to North America in a paper at the 1972 RSNA annual meeting by Ambrose and Hounsfield (1). Those images of the brain not only allowed us for the first time to actually see the human brain noninvasively but also were the first clinical images that were completely dependent on the computer reconstructing a digital image. Since then, the digitization of all types of medical images has rapidly followed, with plain radiographs, computed tomographic (CT) scans, images from nuclear studies, magnetic resonance (MR) images, ultrasonographic images, and, most recently, even high-resolution mammographic images becoming digital in format.
The traditional chest radiograph nicely illustrates some of the advantages deriving from the digital revolution. The traditional chest radiograph is an analog dinosaur created by an x-ray tube that has changed little in the past 30 years and an analog screen-film detector that will disappear. The analog chest radiograph is interpreted by a human observer who will always be important but who will change in function. The chest radiograph is an analog image and is very fragile. I cannot easily reproduce such an analog image, particularly not in a digital format. If you wish to see a chest radiograph that I have, you must either come where I am or I must tediously copy the image at a cost of approximately $4.00 for the 64 megabits of data that it contains. Then I must mail the film to you. The chest radiograph is the product of a manual process that has been slow and has yielded fragile data of moderate information density that have been interpreted in an empirical, qualitative fashion.
The immediate successor to the analog chest radiograph has been the digital chest radiograph. Since this image is originally in a digital format, I can easily reproduce it digitally and easily send it to any computer in my health system network or to you through the Internet (with appropriate security precautions), obviating the costs and logistic challenges of the old analog system. This electronic advance involves use of the same x-ray tube and a new digital detector but the same empirical human observer. It is a digital signal but is forced, for practical reasons, into a hard copy analog image for routine use. These digital/analog images are an improvement, being semiautomatic, faster to produce, and less fragile. However, they are still interpreted in the traditional empirical qualitative fashion.
COMPUTER-ASSISTED DETECTION AND DIAGNOSIS
The digital chest image is where many of us are at the present time. This digital tool uses the same x-ray tube, a new digital detector, and the same human observer, but the image is now displayed as a digital image on computer workstations. This is a much more powerful tool, being faster and more robust. The greater information density contained within the images can now be displayed at the workstation that allows us to better visualize different structures such as bone and surgical clips. Software residing on the workstation immediately and easily allows quantitative measurements, which will become ever more sophisticated.
For instance, in patients with multiple sclerosis, software tools allow one to classify brain tissue as multiple sclerosis lesion, normal gray matter, normal white matter, or cerebrospinal fluid and more precisely measure the volumes of these tissues. The literature is replete with examples that document the superiority of these computer tools to the naked human eye in evaluating the extent and progression of disease (2).
Statistical support for one of radiology’s favorite phrases, normal for age, will be provided by population-based atlases of normality. Statistical atlases of normal ventricles of older subjects have been created (3). An individual’s examination results can then be directly compared by the computer with the normal atlas for that person’s age, and normal for age will then have scientific merit (Fig 1). Even more sophisticated evaluation of an individual’s pattern of cerebral atrophy will assist in the early diagnosis of dementia (4).
|
Computer applications for our digital images will not only be used for more precise quantitative measurements and more powerful visualization. They will also definitely be used to aid us in the detection of disease. Computer-aided diagnosis will become an integral part of our image interpretation. At the 2002 RSNA annual meeting, there were approximately 142 papers on computer-assisted detection or diagnosis and at least 12 new Food and Drug Administration–approved computer-assisted detection or diagnosis packages being demonstrated. Software for the detection of pulmonary nodules on x-ray CT scans and programs that not only render the colon in 3D but also portray in color those regions suspicious for polyp point to the future (5,6). These are among the first of many such tools that will be available for our clinical use. Recent literature strongly suggests that tools such as these will improve the reproducibility, if not the accuracy, of lesion detection and will increase our clinical efficiency. These tools will be at our fingertips on workstations.
IMAGE-GUIDED THERAPY
The use of advanced digital workstations will not just relate to diagnosis, but will increasingly allow the use of our medical images for direct management of therapy. Radiation therapy has been the first and most widely used image-guided therapeutic modality and well illustrates the need for increasingly sophisticated computer analysis. When I was in training back in the 1960s, radiation therapy for carcinoma of the prostate did not require sophisticated imaging, as the technology of radiation therapy itself was limited to large treatment ports that could be guided by crude images. However, radiation therapy technology has dramatically advanced, and more highly localized radiation therapy techniques, including the recently developed intensity-modulated radiation therapy, require very precise 3D images of the target lesion for therapy planning and beam direction.
Image-guided surgery is following the field of radiation therapy. Image-guided robotic devices that use virtual images combined with digital surgical devices enable the surgeon to be directly guided by images. In many institutions today, patients with brain tumors are having image data processed by computers to delineate the tumor area designated for resection. The 3D virtual image of normal tissue and pathologic lesion, digitally linked to a surgical device, is then used by the surgeon for guidance of craniotomy, cortical incision, and tumor resection (Fig 2). It is imperative that the radiologist remain at the lead in the development and even implementation of these image-guided therapies that are dependent on the increasingly complex images we can make and their subsequent computer processing.
|
The complete digital environment, particularly as it is incorporated into an integrated health care enterprise, involves more than digital images and the workstations with their software. The wealth of information in these front-end components of the digital environment must be integrated into the rest of the enterprise. This obviously involves the picture archiving and communication system (PACS), radiology information system (RIS), and equally, if not more importantly, the hospital information system (HIS).
PACS is the intermediary between the inside world of radiology and the outside world of the health care enterprise. The PACS is designed to take the data from the imaging sources and postprocessing workstations and deliver that information through high-speed networks to the archival component of the PACS and other viewing and processing workstations. In addition, it sends and receives information from the RIS and the HIS. Of great importance is easy access to and from the immediate health care system and the rest of the world, particularly patients. This outside communication will mostly be through the Web and must be carefully done to maintain security.
There are many PACS benefits, but the main one is to deliver images and reports to the right place at the right time (Fig 3). For instance, CT images can automatically and simultaneously be routed to the radiologist’s reading station, to the referring physician’s office during working hours, or to the on-call radiologist’s home in the evening. The implications of images at the right place at the right time should not be underestimated. If Dr Smith, the local radiologist, is not immediately available, the images might be sent to a remote reading site. It is inevitable that this technology will result in remote and consolidated reading sites, a feature that offers enormous opportunities in terms of clinical efficiencies but equally enormous management challenges.
|
|
|
clinical correlation suggested. All of a patient’s clinical information will be at our fingertips. In addition, valuable reference material will be within a few keystrokes.
For instance, I recently saw a case of a right cerebellar pontine angle lesion that did not match one of the usual diagnoses but reminded me of a rare lesion—endolymphatic sac tumor. However, I vaguely recalled that this lesion only occurs in conjunction with von Hippel-Lindau syndrome, of which this patient had no stigmata. Wanting more information, I opened the Web browser on my workstation, clicked on my link to the RSNA Index to Imaging Literature, searched on the phrase endolymphatic sac tumor, and retrieved an article on this topic that corrected my memory by reporting that some of these tumors are spontaneous and nonsyndromic. I cut and pasted the reference into my report. Such on time diagnostic assistance not only will improve our work but also allows a very effective learning experience.
The advantages of digital imaging are overwhelming. The process is efficient, clinically powerful, and scientifically critical. The cons, in theory, I believe, are none. In practice, however, there do remain challenges of technologic implementation, cost, and inertia. However, I do believe that contemporary technology now makes this evolution from the analog to the digital environment highly practical and affordable. For example, in our institution, we decided to expand our initial, limited PACS to an enterprise-wide operation. The goal of the project was to convert to a completely filmless operation and to do so on a financial model based on true cost savings.
The costs of an electronic imaging environment have been a significant impediment to implementation, but I would like to remind you of the enormous cost that we have been incurring with our conventional analog film and file room systems. The following are rough figures of our system’s true costs for our film management system. We perform approximately 650,000 examinations per year and spent, in the year 2001, more than $5 million supporting an analog film environment. The cost projected by our finance department over the next 8 years is almost $65 million for this antiquated system. This is an enormous amount of money that I do not believe is out of line with that at many other institutions.
One way of approaching the problem (and one that illustrates real cost savings) is to simply outsource all image management to a commercial partner. Such an arrangement covers the total costs of image management and makes both the purchaser and the provider derive a real contract including all associated costs. In our case, the contract covering all image management costs for 8 years has been accomplished at a cost of approximately $55 million. This includes the cost of all PACS equipment, service, support, and upgrades. It includes all the staffing and management needs for the digital imaging operation. It includes the cost of residual film operations. It will result in true dollar savings to my institution of approximately $10 million over 8 years. This is but one example and one financial approach of many that are now available and which document that the digital environment will not only provide added efficiency to our practices but will actually save money. Importantly, the logistic advantages and lower costs of the digital imaging environment will probably prove of even more benefit to less affluent and well-developed parts of the world.
IHE AND THE PATIENT
So, let’s review the advantages of this digital rEvolution to the patient, the physician, and the health care system. For too brief a time relative to its importance, I will focus on how the digital environment in an integrated health care enterprise could dramatically change our relationship to the patient to the benefit of their health care. This environment will result in timely and more efficient examinations for the patient that will be much appreciated. Furthermore, I propose that this environment will allow not only prompt but more appropriate access to the product that the patient is purchasing from us. The radiologist’s product is a picture and a report. Interestingly, the Health Insurance Portability and Accountability Act (HIPAA) will allow, if not force, us to use the digital environment to relate in a different fashion to our patients.
A recent advertisement from one of our imaging vendors makes an important point. A physician is seen discussing the case with the patient and his or her family. The legend reads, When it comes to patient care, image matters. I do not think any of us would disagree with that statement. Interestingly, the images that contain the rich data that we have talked about are displayed on this illustration, but they are behind the patient. The patient is not seeing the wonderful images. I strongly argue that it should become routine for us to provide our patients with their images, including the increasingly sophisticated and intuitive postprocessed images that are part of our product.
HIPAA is congressional legislation that we have heard a lot about and that deals with a variety of subjects that are not germane to this presentation. For most of us, the discussion of HIPAA has focused on patient record security. However, I believe that one of the more important aspects of HIPAA is that which relates to patients’ rights. The act explicitly states that patients have the right to review and copy their own medical records (Fig 5). They even have the right to request corrections. I believe that patients will increasingly exercise their rights to their medical records and they will do so in order to become more involved in their own care. Our images and reports are an integral and key part of the patients’ medical records, and we should provide them with that information. I would modify the previously mentioned illustration by moving the images from behind the patient to in front of the patient so that they appreciate the role and importance of these images in their own health care. The images are extraordinarily powerful means of communicating, and we should take advantage of this opportunity.
|
a doctor’s doctor for the referring physicians. There will be a variety of means to convey our product to our patients. The product will, of course, be in a digital format. Currently this is quite feasible with compact disks (CDs), onto which images and reports can automatically be recorded. Patients can then take the CD home to their own computer or to their primary care physician or wherever they choose.
Digital advantages to physicians are obvious. The radiologist, after adjusting to the changes, will be very happy having immediate access to all images and ancillary data plus the ability to take advantage of computer-assisted diagnosis and the generation of rapid reports that can be distributed anywhere at any time. Our referring physicians will be ecstatic to have prompt access to our images and our reports. The health care system will be happy because its patients are happy, its physicians are happy, and its costs are less.
So, on behalf of the RSNA, I again thank those many volunteers who have participated in the ECC and its IHE initiative. Their efforts have resulted in radiology being recognized as a leader in the field of medical informatics. I encourage all of us to vigorously and enthusiastically accept this digital rEvolution in our practices.
FOOTNOTES
Abbreviations: ECC = Electronic Communications Committee, HIPAA = Health Insurance Portability and Accountability Act, HIS = hospital information system, IHE = Integrating the Healthcare Enterprise, PACS = picture archiving and communication system, RIS = radiology information system, RSNA = Radiological Society of North America, 3D = three-dimensional
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
