Josep Fernandez-Bayo, Presenter: Nothing to Disclose

Carles Rubies BA, Abstract Co-Author: Nothing to Disclose

Octavio Barbero, Abstract Co-Author: Nothing to Disclose

Melcior Sentis, Abstract Co-Author: Nothing to Disclose

Joan Guanyabens, Abstract Co-Author: Nothing to Disclose

Lluis Donoso, Abstract Co-Author: Nothing to Disclose

Digital mammography remains a challenge due to its high network and storage capacity requirements. It also requires high resolution display monitors. But we think that technology is mature enough to fulfill these high requirements.Our mammography department performs around 30000 studies per year. Considering that a digital breast study has 100MB, we are producing about 3TB per year. We need online access to these studies for at least 5 years, which means an increment of 15TB in online storage capacity in our PACS.The mammography department performs around 120 studies per day with its respective diagnoses. Since radiologists always want to compare with previous studies, we are moving at least 36GB through the network every day. This supposes a considerable increment in network traffic.We use pre-fetching techniques during the night prior to the examination day to minimize this increment on network traffic. Since mammography studies are almost always pre-scheduled, we achieve a significant reduction in network traffic during daylight hours.To minimize the online storage capacity we take advantage of the characteristics of mammography studies: Only 2.5% of mammograms have suspicious malignant findings that have to remain online for at least 5 years. The rest are mammograms with no lesions or short follow up lesions that can be offline until reviewed one or two years later. Thus, theoretically we could reduce the amount of online storage capacity needed from 15 to 0.5TB, but then we have to manually load at least 120 studies from the offline to the online storage every day. In practice we only use 5TB increment online capacity and we have algorithms on the PACS archive that try to minimize the manual load. These algorithms use the integration with the RIS to know the future scheduled exams and to ensure that these exams will be online and ready for pre-fetching.An important issue to address is the coexistence of the new digital system with the previous analog film scenario. It is also important to use lossy compression to distribute these studies to the clinicians. Additionally by using and storing digital mammography, CAD systems can be used.

To describe our experience in digital mammography.To show how we minimize increases in network traffic and storage requirements.To demonstrate that mammography is not an impediment to a fully filmless hospital.

Fernandez-Bayo, J, Rubies, C, Barbero, O, Sentis, M, Guanyabens, J, Donoso, L, Digital Mammography: No Longer the Missing Link to a Fully Filmless Hospital.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4409524.html