Plans to digitize radiology departments often exclude mammography. But advances in digital acquisition, viewing, and network technologies mean that this should no longer be the case, according to a group of Spanish researchers.
Plans to digitize radiology departments often exclude mammography. But advances in digital acquisition, viewing, and network technologies mean that this should no longer be the case, according to a group of Spanish researchers.
Commercial computed radiography and digital radiography solutions for digital breast imaging are already available, said Josep Fernandez-Bayo, a member of the Digital Medical Imaging Center at UDIAT Diagnostic Center in Sabadell, Spain. CR systems offer the more economical route to digital mammography, although DR equipment yields better contrast resolution, and hence image quality, while exposing patients to lower levels of ionizing radiation.
The UDIAT Diagnostic Center chose a DR solution when it decided to convert its mammography services to digital. The center has purchased so-called direct DR equipment that works in combination with a selenium detector, Fernandez-Bayo said. Such systems convert x-ray photons directly into electrical signal, unlike indirect DR systems that rely on the intermediate step of first converting x-rays into light.
The Sabadell center performs about 30,000 mammography examinations each year. With an average file size of 100 MB for each digital study, this equates to an annual data production of 3 TB. Keeping every study accessible for five years would require 15 TB of online storage in the PACS. In reality, Fernandez-Bayo said, only mammograms with suspicious malignant findings need to remain online. This accounts for just 2.5% of the annual workload and reduces the required online storage capacity to 0.5 TB.
Storing mammography data offline, however, creates a substantial load on the network when the studies are reviewed. Technicians would need to manually upload 120 studies from offline storage every day. The department therefore has access to 5 TB online storage and uses its RIS to prefetch offline studies overnight when required. Compression algorithms could also be used to minimize network traffic, Fernandez-Bayo said.
"We can easily achieve 4:1 lossless compression. With this compression, if we still choose the everything-online approach, we can store more than five years of activity in 5 TB. And if we stay with the prefetching approach, we have a significant reduction on manual load," he said.
The existing 100-Mbps network is sufficient to handle either approach. Departments that use multislice CT and mammography intensively would be best advised to upgrade to a 1-Gbps Ethernet, Fernandez-Bayo said.
The department removed its film printer at the end of June, almost immediately after connecting its two new digital mammography units to the PACS. Radiologists now view breast x-rays on dual 5-megapixel monitors. They can arrange the images according to preference, magnify areas of interest, and make precise measurements. Patients who require images for external referral are given digital copies on a CD-ROM.
"There is no sense in acquiring digital mammography equipment only to end up printing studies for storage or reporting. Digital mammography images contain more information than printed films, and the postprocessing that is possible from workstations is part of the added value," Fernandez Bayo said.
Mammography Study Suggests DBT-Based AI May Help Reduce Disparities with Breast Cancer Screening
December 13th 2024New research suggests that AI-powered assessment of digital breast tomosynthesis (DBT) for short-term breast cancer risk may help address racial disparities with detection and shortcomings of traditional mammography in women with dense breasts.
FDA Clears Updated AI Platform for Digital Breast Tomosynthesis
November 12th 2024Employing advanced deep learning convolutional neural networks, ProFound Detection Version 4.0 reportedly offers a 50 percent improvement in detecting cancer in dense breasts in comparison to the previous version of the software.