Integrated IS Speeds Workflow, Delivers Quality Care in Radiation Oncology

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
HIIT070805In the wake of an explosion of oncology-related imaging, many hospitals have now set up separate radiation oncology information systems that not only store images but also keep track of dosage, planning and treatment data.

Traditionally, PACS have only stored medical images. But the kind of medical objects that radiation oncologists deal with are more than just images; they also include dose distributions, treatment summaries and details of the treatment delivery.

The original DICOM specification was for images only; however, the standard was designed so that it could be extended for other data objects. One of the first extensions, DICOM-RT, was designed to accommodate radiation therapy—it was adapted to store information about doses, plans and anatomical outlines.

The implementation of DICOM-RT and imaging-based simulation tools, using modalities such as CT and PET/CT, has eliminated many of the manual inefficiencies that used to plague radiation oncology departments.

Sam Brain, PhD, a senior research associate in the department of radiation oncology at Stanford University’s School of Medicine in Stanford, Calif., says CT scans ordered by radiation oncologists are kept in-house and stored on the department’s radiation oncology information system (ROIS).

Automation integration
According to Brain, Stanford’s Aria (Varian Medical Systems) ROIS stores the multiple parameters required for each patient’s scan: gantry angle, jaw settings and table settings.

“There’s maybe a dozen or so knobs to be twirled before you actually deliver the treatment to the patient. So if you’re doing 100 patients per day and each patient has 30 fractions, that’s a lot of knob twirling to be done, and people make mistakes,” says Brain.

“One of the first things they did when they computerized this process was set up a Record and Verify system,” he says. “The first time you set up a patient it would record the knob settings, and subsequently when you gave the next fraction of treatment, it would check that you set the knobs correctly.”

Stanford’s ROIS also can control multileaf collimators, which aids radiation oncologists in delivering the proper accelerator treatment.

“People started using the computer system for more things,” Brain observes. “It has now morphed into a complete ROIS where we schedule on it, store the CTs in it and store all the anatomical information.”

CT images are used to outline the internal structures of the body, locate tumors and help determine treatment volumes. Prior to the automation of this function, radiation oncologists were drew their treatment contours and calculate treatment fractions by hand. With Varian’s treatment planning software, these processes are optimized at the workstation and the treatment details are stored in the ROIS.

All radiation oncology information is stored on a central server, which contains medical images that can be imported from the radiology department, including details of internal anatomy, dose details, treatment plans and galleries of the radiation of the tumor. Radiation oncologists look at these so-called “portal” images to determine if they are “hitting” the tumor, Brain says.

Stanford’s Aria system has greatly helped to unite radiation oncology care, he asserts. “It means that a physician can go to a terminal anywhere in the dept, sometimes from home or on the road, and by using properly protected communications protocols, can check some details of the [treatment plan for a] patient,” Brain notes. “So the fact that it’s all in one place is fantastic and has really helped our workflow.”

It’s a big improvement over former piecemeal systems made my different vendors, he says. Aria is able to communicate with other systems in the department—such as its PET/CT simulator or CT simulator—and pull out the details needed to make treatment calculations, Brain says.

Clinical flows are quite complicated, Brain says, and involve scanning, outlining, planning and treating. Data flow is a multistep process: CT images are fed into the central server; the outlining workstation pulls those images out and a specialist will outline tumor margins; this data are then saved on the central server; the planning system then pulls the images and outlined tumor margins and a plan is calculated by an oncologist; finally, at the accelerator, the plan is downloaded and used for treatment.

“Everything to and from a central server is the optimal way of doing things,” states Brain, adding that having all components from one vendor eliminates data translation or DICOM-compatibility problems.

Multispecialty reach
George Laramore, MD, PhD, a Peter Wootton professor and chair of the department of radiation oncology at the University of Washington Medical Center in Seattle, says oncologists, radiologists, surgeons, multidisciplinary tumor boards, as well as the physicians who individually see patients, uses their Elekta Impac ROIS.

The system is used for diagnosis, to locate tumors, for staging purposes, to look at possible interventions and then decide which are preferable in terms of tumor control and side effects, Laramore says. In addition, radiation oncologists use it for treatment planning.

The centralized information system unites the oncology care team, he says. The Impac ROIS allows access to diagnostic images stored on the facility’s GE Healthcare Centricity PACS. “You simply pull up a patient’s name and you see what images are stored there,” Laramore says.

The ROIS-PACS integration works seamlessly except when the images are not taken in-house, and patients come in with CDs, which take a few days to upload into the PACS, he notes.

The ROIS serves multiple hospitals: the Seattle Cancer Care Alliance; the Harborview Medical Center; the University of Washington Medical Center; and all nine of the University of Washington primary care physician network sites that can access it.
All of the facilities have access to the same PACS; but for radiation oncology, the portal images are stored on an Impac data server that’s not integrated with the PACS.

“It’s a separate system, not for diagnosis, but for validation of daily treatments,” Laramore explains.

Both Laramore and Brain note that their ROIS has a similar issue: there’s too much data to manage.

“In most cases, I don’t need all the information I’m getting on the images,” says Laramore. “What I want to do is use the images to validate that I’m treating the volume that I want. If I can do that validation without acquiring all these gigabytes of data, I’d be happy because I can’t store them. I’m never going to look at them. So we need to have some way of filtering the information so it gives the clinician the answer to the question that’s being posed.”

Observes Brain, “For me, as the IT guy, it means the image sets have become 10 times bigger, and the amount of information we’re having to store in our server is exploding, and you can’t throw any away because maybe the MD wants to look at it later.”

The most valuable commodity in a radiation therapy practice is the radiation oncologist’s time. As patient volume continues to increase, technology solutions such as a ROIS offer a department a tool for ensuring the efficient and effective allocation of their most precious resource.