On the final day of SIIM 2015, a vendor panel took place that looked at the evolution and adoption of digital breast tomosynthesis (DBT), and whether or not facilities were prepared to handle the influx of these 3D images on their PACS, as well as across their IT networks.
The panel was led by David A. Clunie, PixelMed, and consisted of Ron Muscosky, MSEE, Carestream, Steve Deaton, Viztek, Mark Bronkalla, MBA, Merge, and Bobby Roe, Visage.
After the vendors each took turn describing their own offerings and advice to those who currently use DBT or are planning to, we segued into an audience Q&A that dived into more detail among the seven audience questions. Below are the questions asked, as well as answers provided from the panel.
Question #1: Secondary capture–should facilities be talked out of it?
Answer: The option exists to store in a standalone archive, or in a PACS and then have the DBT images converted. The issue is that once converted into a PACS, facilities now have double the storage needs as two images now exist. Among panel attendees, they all either had DBT in place, and planned to have it within the next 12 months. However, few who were planning deplpy DBT had a plan in place. Essentially, facilities need to put priority on BTO (Breast Tomosynthesis Object) over SCO (Secondary Capture Object) for improved flexibility in storing and sharing images.
Question #2: What is the clinical impact of DBT? Does it take three times as long to read the exams?
Answer: This answer to this question was diverted to the audience, where one attendees traditional 2D mammography took 10-20 seconds to read, versus DBT exams which took 20 minutes. Attendees knew reading DBT exams took longer, but not that long. The panel responded to this answer by saying that DBT was still relatively new and as adoption and familiarity increased, reading time would become more efficient.
Question #3: Will reimbursement for DBT improve? Especially after the stick-shock for installation?
Answer: When DBT was not reimbursed, few were using the technology. Now that more are using the technology, reimbursement is in place, though it is consider poor. For improved reimbursement, the introduction of new competitors will drive purchase and installation prices. Carestream’s Muscosky said that he has been seeing many facilities adding DBT for competitive reasons, because they do not want to lose patients to other facilities that have it installed. There are of course diagnostic benefits to DBT, and in turn providing better patient care. Essentially, facilities should be installing DBT to maintain an edge on the competition, and use it as a marketing device for the facility. A large majority of the audience agreed with this and are currently practicing this themselves.
Question#4: Regarding size of images–how can DICOMweb help with performance issues?
Answer: One response was that is does not matter one way or the other if facility is handling everything on the server side. Another response was that any image being view for mammography is either lossless or not. High resolution, no bandwidth advantage to the client, because you want to have a viewer that does not need the entire study to start display or the entire object to start display. Server-side issues have advantages but is not a silver bullet. Facilities can user server-side for non-diagnostic, or offer a choice on whether facilities want loss or lossless, and server-side is not truly offering bandwidth advantages.
Question #5: This audience member asked about digital mammography priors for comparison reasons that has resulted in a struggling workflow. She had to alter workflow because her facility cannot get to prior to workstations fast enough. Has this problem been fixed with digital mammography?
Answer: The issue with the attendee’s facility was that there are multiple offices, but read at one central workstation. The facility is sending all studies to all workstation, creating three or four copies. Routers would send study once across, and central location would route once. The solution to this is that intelligent routing would mitigate the issue. Technology that is zero-footprint has not been available, whether supporting server-side rendering or not, but it is moving in that direction. All of this re-routing goes away once facilities move to the cloud. Panelists commented how it is important to focus on not just getting diagnostic tools to diagnosing physician, but also having good bedside manner with patients. Carestream’s Muscosky added that many facilities are converging multiple sites, and data needs to be accessed quickly no matter where from. Data cannot be sent around among facilities and re-routed.
Question #6: Is the industry moving toward supporting motion detection and correction?
Answer: Panelists all agreed that this advancement would need to be completed on the image acquisition side, since these vendors are the ones creating the hardware. They all agreed that it would great to see projection when there is motion, but have not heard of anyone released a motion detector.
Question #7: How do facilities correct hanging protocols that are not working?
Answer: One true way to identify an image according to view, and every view port has to give the user an option to toggle between C-view, and others. Facilities should refer to the IHE DBT framework. There was an idea presented that in the future, vendors should allow hanging protocols that make more sense and are deterministic. Users still want to hang right, regardless of vendor, and new modality gives a chance to hope for the best compliance. In the future, there will need to be a right code that allows users to apply logic and alternate paths to identify mishaps.
Rich Pulvino is the digital media specialist for Carestream. He attended SIIM 2015 from May 28-30 at the National Harbor in Maryland.