Comparison of ion chamber and TLD dosimetry in mammography

An ionization chamber method has been developed to measure exposure vs depth in a uniform BR 12 "average breast" phantom. It employs a Memorial mammography chamber for exit exposure measurements; resulting data is then corrected for backscatter as well as for the exceptionally thin window of this chamber. A careful comparison has then been made with relative exposure vs depth curves obtained using TLD at several mammography beam qualities, for identical exposure factors and SSD values. Use of a correction for residual and background TL signals significantly improved agreement between TLD and ion chamber curves in the 28 to 35 kVp/0.03 mm Mo range of beam quality. Agreement was within +/- 5% for the Mo target tube, but TLD readings were 4%--8% higher than ion chamber values for the W/Mo target tube. At Xeromammography energies (45 kVp/1.6 mm Al), corrected TLD curve readings were 6% higher at depth than ion chamber curve values. TLD meaurements with 28 to 35 kVp/0.03 mm Mo beams tend to underestimate dosage to the midbreast parenchyma. For example, in a 5 cm "average breast", the underestimation ranges from 2%--10% for corrected, 10%--16% for uncorrected TLD readings.