Reducing radiation dose from traditional mammography in women with breast implants was the subject of an AuntMinnie.com article earlier this week. Researchers from California suggested using fewer views and a faster target-filter combination to reduce dose.
On a similar topic, investigators at Duke University looked at the possibility of reducing dose in full-field digital mammography (FFDM) and whether less radiation exposure would hamper diagnostic performance.
In theory, transitioning from analog to digital should offer an opportunity to reduce mammographic dose, explained Ehsan Samei, Ph.D., and colleagues. But how a lower dosage would affect image quality remains unclear, resulting in the same dose for film-screen mammography (FSM) and FFDM in the clinical setting.
"The purpose of our study was to experimentally determine the relationship between radiation dose and observer accuracy in the detection and discrimination of simulated lesions for digital mammography," wrote Samei's group, which is from the departments of radiology, breast imaging, physics, biomedical engineering, and biostatistics at Duke University Medical Center in Durham, NC (Radiology, May 2007, Vol. 243:2, pp. 396-404).
For this research, 300 normal craniocaudal images were selected from a database. All images were acquired on an indirect flat-panel mammography system (Senographe, GE Healthcare, Chalfont St. Giles, U.K.) with an average tube voltage of 27.6 kVp.
The average compressed breast thickness was 5.1 cm, and densities ranged from fatty to extremely dense. Simulated mammographic lesions included masses (benign and malignant) and microcalcifications, divided among the 300 images.
"To create images with an image noise similar to that caused by a reduction in radiation dose, a noise modification routine was used to add radiographic noise to the images," the authors stated, so that the images were place into three subgroups: full dose (no added noise), half dose, and quarter dose. Postprocessing included unsharp masking and contrast equalization, as well as window and level settings processing.
Five experienced breast imagers viewed the scans on a 5-megapixel LCD monitor (Nova V, National Display Systems, Morgan Hill, CA) and used a "location-known-exactly" experimental paradigm to score the images. Each observer took about two and a half hours to read the entire set of images.
According to the results, there was a reduction in overall accuracy that corresponded with reduced radiation dose. This was particularly notable for the transition from full dose to quarter dose, although the reduction was less marked when going from full dose to half dose, the authors stated. They also saw a "clear decrease" in the detection of microcalcifications with the reduce dose, as well as a decrease in mass discrimination.
Finally, for a half dose, the median interpretation time per image was 2.42 seconds, which increased to 3.04 seconds for quarter-dose images. The median interpretation time per image was 2.38 seconds for those obtained at full dose.
"We found that decreasing the dose in digital mammography by as much as half had a minimal effect on the detection of malignant masses but a notable effect on the detection of microcalcifications, the discrimination between benign and malignant masses, and the interpretation time," the group concluded.
They acknowledged that this study had some limitations: the study did not look at the quantitative relationship between glandular breast dose and accuracy, and it was a simulated study.
However, as FFDM dose research moves forward, the results of this study did show that, even with FFDM's higher detective quantum efficiency, dose reduction has a measurable -- albeit moderate -- affect on diagnostic accuracy, the authors added.
By Shalmali Pal
AuntMinnie.com staff writer
May 11, 2007
Related Reading
NEMA's MITA touts dose-reduction technologies, April 25, 2007
RSNA studies delve deeper into DMIST results, December 14, 2006
More DMIST analysis supports FFDM in younger women, dense breasts, November 26, 2006
Copyright © 2007 AuntMinnie.com
