(Saturday, 4/4/2020) 10:30 AM - 12:30 PM [Mountain Time (GMT-6)]
Purpose: Multiple Hologic Selenia Dimensions mammography units were identified as having a fixed offset of the measured peak tube voltage (kVp) during digital breast tomosynthesis (DBT) exposures but not during full field digital mammography (FFDM) exposures. At lower kVp stations this offset caused the kVp accuracy measurements, defined as the percent difference between the indicated kVp and measured kVp, to exceed the manufacturer’s limit of 5%. The root cause was investigated and identified.
Methods: The kVp accuracy was measured using a RaySafe X2 for both FFDM and DBT acquisitions to determine the magnitude of the offset. Invasive measurements of the peak kVp demonstrated no offset. Working with Hologic, kVp as a function of time was measured using an oscilloscope. The waveform was modified and kVp accuracy was re-measured. A historical survey of kVp measurements for 16 units was performed to determine the prevalence of the issue.
Results: The kVp waveform’s leading edge contained an overshoot and was identified as the cause. After flattening the kVp waveform and resetting the kVp, the kVp accuracy was brought within 3.5% of all measured stations. The measured average glandular dose (AGD) for DBT acquisitions following corrective action was reduced from 1.48 mGy to 1.25 mGy (p < 0.001) on one affected unit. All image quality metrics were maintained at their pre-adjustment levels. A total of 3 out of 16 units were identified as having this offset.
Conclusion: A systematic offset of the measured kVp present only in DBT exposures was identified. The root cause was found to be an overshoot present in the kVp waveform that did not impact FFDM measurements. Corrective action resulted in a lower AGD for DBT exposures while maintaining FFDM AGD and image quality metrics. This offset may be present in other machines in clinical use.
Contact Email
