Follow the Leader

Eliot Siegel reveals to EHM how he and his team revolutionised radiology and made film a thing of the past.

Transforming imaging informatics

“We've moved to a digital environment and patients are now being handed CDs or DVDs after they have their CT or MRI studies”
-Eliot Siegel, Professor and Vice Chairman of the Univeristy of Maryland School of Medicine

In 1993 Eliot Siegel, Professor and Vice Chairman of Information Systems, and his partners at the University of Maryland School of Radiology, were the first radiologists to introduce film-less technology. Fifteen years on and they have continued in their innovative quest to transform imaging informatics, producing technologies ahead of the digital age.

“I maging informatics can be thought of in diagnostic imaging as a subset of medical informatics, the field of study concerned with the broad issues, management and use of biomedical information, including the study of medical information,” explains Siegel. He notes that imaging informatics is defined as the subset of medical informatics, which touches on every aspect of the imaging chain

“That includes not only the creation and acquisition, distribution and management of images, their storage and retrieval, but also imaging processing, image analysis and image and navigation, and image interpretation and reporting and communication and many other areas,” he adds. “Imaging informatics is the nexus between diagnostic imaging and other disciplines, including engineering, information, technology and physics.”

Challenges

Being at the forefront and creating such innovative technology brings with it many challenges. “One of the particularly interesting areas creating media attention recently has been the optimal trade-off with regard to dose and image quality central to diagnostic imaging,” Siegel says. “What is the definition of image quality and how can we actually measure it and improve it? Is image quality just defined as what is aesthetically pleasing to the radiologist, or is there a more general quantifiable definition of it?”

Quantification does not come without its difficulties. Its function in diagnostic radiology is to provide tangible results through enumerate means. “When using CT, MRI or other modalities as a metric for patient change, we need to have more rigor in the way that we measure lesions and in our criteria for determining size, volume or what is the error of measurement,” Siegel says. Determining these results on a quantifiable basis allows for diagnostic radiology to move into an era of personalized medicine.

“Quantitative diagnostic radiology provides the ability to use the patients individual DNA and the tumour’s DNA and correlate that with laboratory and quantitative diagnostic radiology information, and through making all of those fit together we can tailor a specific treatment or screening regime for a particular patient.”

Quantifying results also produces the benefit of greater communication between patient and physician. “An important role for imaging informatics is ensuring that this information is communicated properly to the physicians taking care of the patient, and also that there is acknowledgement back from those physicians that they’ve received the message,” Siegel points out, adding, “From this, we’re able to track whether or not recommendations that we’ve made are actually followed up.”

Another important imaging informatics challenge is to ascertain the trade-off between productivity and diagnostic accuracy. “So many of us in diagnostic imaging are starting to feel like I Love Lucy’s Lucy and Ethel from in the candy factory, where there’s an assembly line and the complexity of studies and volume of studies is increasing. How can we use information technologies to be able to increase our diagnostic accuracy while increasing productivity? Sort of a paradoxical increase in both of those would be really wonderful, and it’s one of the things that we investigate,” he explains.

Innovation

It is not only diagnosing an accurate interpretation of informatics imaging that poses a problem; communication with physicians brings with it technological challenges as well. Siegel and his team faced those difficulties 15 years ago when they unveiled their creative innovation and introduced film-less technology.

“We were the only department that was film-less in the United States for quite a few months and in order to interface with our incoming patients, along with being able to share our images for patients who were seen at other hospitals or clinics in addition to ours, we had to resort to interfacing using film,” Siegel recalls.

As other facilities have made their transition towards digital imaging, things have not necessarily become easier. “Although we’ve moved to a digital environment and patients are now being handed CDs or DVDs after they have their CT or MRI studies, the problem we have now is a ‘Tower of Babel’ situation of confusion due to the different formats which those CDs are written,” says Siegel. In an era of vast technological advancements, the communication of images from one facility to another is made harder due to a lack of standards that exist for the interchanging of information even using standard media such as CDs.

“I would see the transition in CDs developing in the future towards a direct electronic mechanism that allows me to access my information, regardless of the hospital I’m in, in a manner analogous to when I go to an ATM machine to access my US account when I visit London. We’re investigating the standards for the capability to be able to electronically exchange that medical information in a safe and secure way. This is being done through a project called IHE – Integrating the Healthcare Enterprise – which is a joint effort between the Radiologic Society of North America and a computer information technology site called IMSS.” Siegel believes technological development is progressing toward a direct electronic mechanism in which this incompatibility of formats can be corrected.

Within Siegel’s own department, the changing nature of imaging informatics is exemplified by the innovations currently being created. “We’ve radically redesigned our radiology reading room, and created what we call the radiology reading room of the future, which embraces all of these technological challenges,” he explains. “Many of the institutions that have made the transition from film-based radiology to film-less have merely substituted computer workstations for the viewing boxes without thinking of changes required in lighting, ergonomics and seating. We’ve done a lot of work in working with architects and experts to completely redesign our radiology reading room.”

Technologies

The department has also introduced speech recognition technology, which, with the elimination of the transcription process, allows for the ability to decrease report turnaround times. Advanced 3D workstations are a major development from the film-less technology that was previously used, creating a much shorter timeframe in which images can be received.

“Fifteen years ago, we were looking at images electronically in a much more passive way, whereas now we’re navigating through 3D space with advanced visualization systems. We’re interacting as radiologists and determining the way we want to look at the images, rather than the way the patient fits into the CT scanner,” Siegel says. He compares the technology being used within the department as similar to that of Google Earth. The Google mechanism of looking at maps is translated within informatics to communicate information via a server that is able to provide advanced imaging processing and visualization.

“The ultimate affect of this on the patients is that they can now come to our department, and without having to have additional sub-specialized studies, routine studies can now be reconstructed so that we can get very detailed views of the spine or the patient’s vasculature, pulmonary vasculature or abdominal vasculature. During one visit, we’re able to acquire information volumetrically, and the benefit of this for the patient is that we’re able to make more rapid and more accurate diagnoses using less intravenous contrast than we were previously.”

The innovative technologies used by Siegel and his team are applied to each of the 30 to 40 projects he may be working on at any one time. In response to progressing quantifiable measurable results, Siegel explains the work he is doing on algorithms: “We’re investigating different types of algorithms and ways in which to make better volumetric quantitative measurements on patient lung lesions, rather than just making axial or coronal measurements.” Siegel and colleagues are investigating the use of grid computing which can facilitate the ability for multiple computers to work on an imaging challenge such as the detection of lung nodules in a patient to either decrease the time required for computer assisted diagnosis or create a consensus among multiple different algorithms working in parallel. Siegel adds, “We can also use grid technology to move the algorithms to the diagnostic imaging center without having to move the images themselves.”

“We’re also conducting an ergonomic study, evaluating the impact on diagnostic accuracy and the physiologic impact on radiologists walking slowly, somewhere around one mile per hour, on a treadmill while doing image interpretation. This will help us measure the physiologic impact and other impacts on radiologists,” says Siegel. “We physicians tend to take better care of our patients than we do ourselves.”

Technological innovation is what propels Siegel and his team to the forefront of discovery. One project he is currently working on focuses on the development of multi-touch technologies, and the impact such mechanisms can provide within radiology. “The ability for a radiologist to be able to navigate, not with a mouse or a trackball, but via a multi-touch interactive screen when looking at a complex CT or MRI dataset will dramatically change the way information is visualized.”

Discovery

One of the major projects Siegel has just received funding for is the building of a new CT scanner within the department. “We’ll be creating our own scanner technology, using different types of dose detectors in which there will be a significant reduction in the dose of radiation in comparison to what is conventionally used in CT,” he says. “We’re also looking at different ways that radiologists can report findings out, so rather than just using a text method for reporting, we’ve done some research looking at gesture based reporting. The effect of this is the ability of a radiologist to place symbols or marks on an image, and then having the computer generate a report based on that.”

“We’re also looking at the impact of reducing radiation dosage on computer programs that do automated computer-aided detection as well as novel CAD applications such as the creation of computer aided detection programs for some new novel applications, for example evaluating for meniscal or tendon tears within the knee.”

For Siegel and his radiology team at the University of Maryland School of Medicine, becoming pioneers of imaging informatics innovation did not end at the creation of film-less technology. Their research and technological developments have produced intriguing results and this has furthered their desire to remain at the forefront of technological advancements. As Siegel’s current research projects show, there is certainly a buzz of excitement around the anticipation of technological development in imaging informatics technologies in the next 15 years.

Elliot Siegel is Professor of Diagnostic Radiology and Nuclear Medicine at University of Maryland Medical Center.