Optimal contact-pressure for Digital Breast Tomosynthesis

In conventional mammography, one of the reasons for breast compression is
lateral spreading of the tissue to reduce overprojection. Because Digital
Breast Tomosynthesis (DBT) combines multiple projection images from various
angles into a pseudo 3-dimensional image, overprojection is less of a concern.
It has therefore been hypothesised that less compression may be sufficient,
however, all results in literature are based on reducing the target force. A
2015 PhD dissertation from the University of Amsterdam shows that
pressure-standardized compression is preferred because it adjusts the
compression force to the size and firmness of the individual breast. This is
shown to reduce pain experience without compromising the image quality or
increasing the absorbed glandular x-ray dose (AGD). Therefore, the AMC
standard-of-care compression protocol since 2014 is to use a contact area
average pressure of 10 kPa (* 75 mmHg) for all women. For DBT, it is however
not yet known which level of contact-pressure (compression force divided by the
breast-paddle contact area) is optimal.

The objective of this study is to statistically identify the optimal level of
contact-pressure for DBT. *Optimal* is defined as a weighed sum of three
clinical interests: as high as possible lesion conspicuity, as low as possible
absorbed glandular x-ray dose and as low as possible pain experience.

The objective is most directly achieved by using a within-women comparison
study. Each breast will get a customized two-in-one DBT-recording whereby two
DBT image sets are acquired at two different levels of contact-pressure without
repositioning the breast. Three radiologists will compare lesion conspicuity,
the absorbed glandular dose is retrieved from the DBT image information header
(DICOM) and the participants are asked to rate their pain experience directly
after each of the two DBT image-acquisitions. The ANOVA test will be used to
select which level of contact-pressure is optimal: (rounded) 6, 8, 10 or 12
kPa.

One of the DBT-recordings per included breast is replaced by a custom *two-
in-one* DBT-recording. Instead of releasing the breast immediately after the
x-ray exposure, the mammography technologist will apply slightly more pressure
and make a second DBT-recording.

Participants will receive one extra DBT-recording per included breast on top of
two routine recordings per breast. The extra DBT-recording will increase the
total absorbed glandular dose per included breast but has the advantage of
providing a second image set in which the breast is exactly in the same
position but slightly more compressed. Differences between these two image sets
will be visualized by blinking, which highlights the deformability of internal
structures. Combined with the pressure-levels, we will calculate the strain
map, which quantifies the elasticity of internal structures. Both can increase
the conspicuity of malignant lesions.

For obtaining the two-in-one DBT-recording, the breast will be flattened and
immobilized approximately twice as long (60 to 90 seconds instead of 30 to 45
seconds). This could be more painful, however, all compression levels are based
on the contact-pressure: the forces are therefore already adjusted to the size
and firmness of the individual breast. The study will add two to five minutes
to a total of 10 to 15 minutes.