Advanced viz tool development aids whole-breast ultrasound efforts
Whole-breast ultrasound offers the potential for CT-like clarity in breast imaging without the load of ionizing radiation while the volume of images from these exams poses a throughput issue for interpreting clinician workflow. Advanced visualization tools can assist with the deployment of whole-breast ultrasound in clinical practice because fatigue from interpreting a large volume of breast ultrasound images can contribute to oversights of masses, according to a team of scientists from Chunichi Hospital and the Nagoya Medical Center in Nagoya, Dokkyo Medical University in Mibu, and Gifu University in Gifu.
They discovered that if an automated recognition system could extract skin, nipple, rib, fat, pectoralis and mammary gland from whole-breast ultrasound images, the results could be applied to a computer-assisted detection system and a computerized registration of other modality images. Their preliminary work proposed methods for extraction of skin and nipples from whole-breast ultrasound images.
High-use flat-panel displays require conscientious QC
“To prevent medical malpractice, quality control (QC) for luminance of high-quality medical displays at PACS terminals is very important,” according to a team of researchers from Nagoya City University Hospital, Kanazawa University School of Health Sciences and Fujita Health University in Japan.
The team investigated the correlation between the frequency of use and luminance degradation in heavily used displays for image interpretation in the department of radiology and the less-frequently-used displays at other departments.
The displays were divided into two groups: the displays at the PACS workstations used for a considerable time each day in the radiology department; and display monitors that were used less frequently.
The average cumulative hours of use for radiology displays were 144,880 hours, while those for other department displays were 2,192 hours.
They observed that when the highest luminance became 350 candelas per meter squared (cd/m2) or less, luminance hardly recovered, even though calibration was carried out. The team’s research found that after about 20,000 hours of cumulative use, luminance had degraded to approximately 400 cd/m2.
THE BUSINESS OF IMAGING
CT: Doing more but getting less; Rads recommend fewer procedures over time
Despite continuing technical advances and increased use of CT in clinical practice, the diagnostic yield of CT exams has actually decreased over the past decade, according to data from a Natural Language Processing program, Leximer, developed at Massachusetts General Hospital in Boston, that was used to analyze reports of all CT exams performed at the institution from 1996 to 2005. Pragya Dang, MBBS, reported that the group found there was a 14 percent growth in volume of CT exams, which outpaced the increase in recommendation rates for further imaging studies in the radiologist’s report (13.6 percent to 19.6 percent, increasing by 0.5 percent per year). “Over this time, the rate of findings in radiology reports changed from 79.5 percent to 75.4 percent, decreasing at a rate of 0.6 percent per year,” Dang said. The researchers found that recommendation rates increased for all radiology specialties; however, finding rates decreased for all radiology sub-specialties, except neuroradiology.
Using the same system, researchers also found that the longer a radiologist is in practice, the lower his or her rate of requests for additional diagnostic imaging procedures. A decrease in recommendation rates was observed with an increase in years of experience of radiologists in abdominal imaging, neuroradiology, thoracic and musculoskeletal radiology.
CPOE and strategies for taming the CT growth beast
Massachusetts General Hospital (MGH) in Boston has contained growth in outpatient CT imaging volume by basing its CT-diet regimen on computerized radiology order entry with decision support. “Computerized radiology order entry with decision support is an effective way to reduce the image intensity creep,” said Dang (see reference above).
MGH introduced web-based computer order entry in 2001 and rolled the system out across the entire outpatient physician population over the next two years. The hospital launched decision support in 2004, requiring ordering physicians to enter indications and demographics for imaging