Quality Assurance and Radiation Dose Monitoring for Digital Mammography Using the Dose Index Tracker
 
Authors:
Mary B. Peter, MS, Mayo Clinic – Arizona; Beth A. Schueler, PhD; William Pavlicek, PhD; Steven G. Langer, PhD; Teresa Wu, PhD; Shanshan Wang; Mengqui Hu, MS
 
Background:
Digital images stored with a Digital Imaging and Communications in Medicine (DICOM) header contain information that can be valuable for quality assurance monitoring. Capture and transfer of this information to a central database allows for automated review of exposure parameters and radiation dose, specific to individual acquisition systems, exam types, or technologists. With data collected from a large number of patient examinations, information about typical clinical exposure parameters can be extracted to guide selection of relevant equipment settings for performance evaluation, for comparison of radiation doses to benchmark reference values, or for monitoring technologist working habits. The Dose Index Tracker (DIT) was designed to perform this task. This information system extracts specified DICOM header field data elements and populates a database for subsequent data analysis.

We have applied DIT to monitor exposure information from a digital mammography unit. This monitoring is used to collect data on x-ray beam energy, anode and filter type used, compressed breast thickness, compression force, entrance skin dose, and average glandular dose (AGD). Results are compared to published thickness and radiation dose measurements for comparable acquisition systems. Results are also used to select appropriate exposure parameters for equipment performance testing of automatic exposure control, x-ray beam energy and compression thickness readout systems, and to monitor compression force levels used by individual technologists.
 
Evaluation:
Data was extracted from mammography images acquired with a General Electric Full-Field Digital Mammography System (Senographe Essential, GE Medical Systems, Milwaukee, WI). In automatic exposure control mode, the imaging system selects the appropriate anode (rhodium (Rh) or molybdenum (Mo)), filter material (Rh or Mo), peak kilovoltage (kVp) and milliamperage-time (mAs) based on compressed breast thickness and attenuation. Data from a population of 100 women receiving bilateral screening mammography examinations are reported in this study. As image acquisition methods for women with breast implants differ from standard automatic exposure control, these images were excluded from the study.

Values for entrance dose and average glandular dose given in the DICOM header were validated by comparing to an ionization chamber measurement. The reported entrance dose (air kerma) value was found to be within < 1% of the measured value. Using the measured half-value layer, actual compressed breast thickness, and 50% glandular breast tissue dose conversion coefficients[1], the average glandular dose value was found to be within 3% of the measured value.

Figure 1

Figure 2

Patient age ranged from 36 to 87, with a mean value of 61. The number of images per exam ranged from 4 to 7, with a mean value of 4.4. Compressed breast thickness per view ranged from 22 to 84 mm, with a mean value of 53 mm (see histogram in Fig 1). Compression force ranged from 30 to 160 newtons (N), with a mean value of 93 N (see histogram in Fig 2). Less than 1% of the exposures were acquired with the Mo/Mo anode/filter combination. The Mo/Rh anode/filter combination accounted for 15.5% of the total exposures, while the vast majority (83.8%) were acquired with Rh/Rh. Entrance dose (air kerma) per view ranged from 2.16 to 10.33 mGy, with a mean value of 5.39 (see histogram in Fig 3). Average glandular dose ranged from 0.73 to 1.85 mGy, with a mean value of 1.20 (see histogram in Fig 4).

Figure 3

Figure 4
 
Discussion:
Compressed breast thickness at our facility was found to be comparable with other published mammography surveys, which report mean values ranging from 49 to 52 mm[2-4]. Average glandular dose was also similar to results from several recent digital mammography dose surveys where mean values ranged from 1.3 to 1.9 mGy per exposure[2,5].

Results of the compressed breast thickness and compression force data were used to assist with selection of parameters for equipment thickness scale readout testing. To correspond to clinical parameters used at our facility, the thickness scale is tested for phantoms ranging from 20-80 mm, with a compression force of 90 N. It is notable that this value is significantly higher than the value recommended by the mammography system manufacturer (50 N) [6]. Similarly, kVp data was used to select settings for kVp accuracy and reproducility testing. Data collected indicates the most commonly used clinical kVp at our facility is 29, with minimum and maximum kVps of 26 and 31, respectively.

Figure 5

Compression force data was also evaluated individually for each technologist (see Fig 5). It was found that Techs 2, 6, and 7 tend to use less compression force than their peers, which may indicate a need for corrective action.

In the future, we plan to extend use of DIT to monitor images from additional mammography units (from different manufacturers). In addition, we plan to include the additional DICOM header elements for monitoring of exposure index, flat panel detector temperature, automatic exposure control mode, compensation step selection, and presence of breast implants.
 
Conclusion:
A method to extract data from digital mammography images has been developed and applied to a patient population. Analysis of these exposure parameters allows for detailed, near real-time monitoring of the clinical practice, resulting in expanded (more encompassing) quality assurance methods, accurate selection of equipment testing parameters, and simplified collection of population dose and image quality information.
 
References:
1. Wu X, Gingold EL, et al. “Normalized average glandular dose in molybdenum target-rhodium filter and rhodium target-rhodium filter mammography.” Radiology. 1994;193(1):83-89.

2. Moran P, Chevalier M, et al. “A survey of patient dose and clinical factors in a full-field digital mammography system.” Radiat Prot Dosimetry. 2005;114(1-3):375-9.

3. Rosenberg RD, Kelsey CA, et al. “Computer-based collection of mammographic exposure data for quality assurance and dosimetry.” Med Phys. 2001;28(8):1546-51.

4. Kruger RL, Schueler BA. “A survey of clinical factors and patient dose in mammography.” Med Phys. 2001;28(7):1449-54.

5. Gennaro G, Baldelli P, et al. “Patient dose in full-field digital mammography: An Italian survey.” Eur Radiol. 2004;14(4):645-52.

6. Senographe Essential Acquisition System Quality Control Manual, Revision 1. GE Healthcare. 2008; 5305863-3-S-1EN.