The retrospective study included 100 women submitted to mammography in our hospital,
randomly matched into two groups according to their age,
body mass index,
breast density category,
and compressed breast thickness: first group was filmed using AEC-device in function,
while the second group consisted of women filmed in the period od AEC malfunction,
in which the exposure parameters were determined manually by mammographer.
Manual exposure parameter control is nowadays mostly abandoned in mammography,
as the AEC unit is able to optimise film exposure relative to image quality in most cases.
Manual technique still might be superior in certain patients such as those with unusual body/breast habitus,
emphasized posttreatment breast inhomogeneity,
breasts unsuitable for proper compression or uncorrectable equipment malfunction.
Two mammographers with profound practical skill in the method (24- to 26-yrs of experience in clinical mammography) were assigned to film the women without AEC-device use,
as the earlier appointments of patients could not be cancelled.
Among eleven radiology technologists (RTs) who perform mammography on daily basis,
only two senior coleagues were sufficiently experienced in manual exposure management,
while the other predominantly younger technicians were not adequately trained due to ubiquitous AEC use („AEC-addicts“) .
Two skilled radiologists with >10 yrs od experience in mammography analyzed image quality according to European Comission (EC) guidelines in both groups.
No significant difference was observed,
with average score of 2,63 ± 0,67 in the group with automatic exposure control use,
and 2,60 ± 0,70 in the group with manual technique.
Mean kappa values for the examined cohorts was 0,74 ± 0,20 and 0,75 ± 0,10 respectively (Table 1 and Table 2).
The reviewing radiologists analyzed each mammography by 9 different categories (brightness,
contrast,
sharpness,
noise,
artefacts and detailed depiction of anatomical structures,
such as skin,
retro-mammary space,
glandular tissue,
and the detection of microcalcifications) using a 4-level ordinal scale (0-not applicable,
1-inadequate,
2-moderate and 3-excellent).
Comparable superior image quality was observed in both groups,
with virtualy no influence by the difference in techniques. The study was not blinded for reviewers,
hence the radiologists might have been biased by the knowledge that highly experienced technologists performed the mammography.
However,
high kappa values and comparable image quality scores reflects achived high image quality in both groups.
The mammography was done using a full flat-panel detector (FFD) mammography scanner (Mammomat Novation DR,
Siemens,
Erlangen,
Germany). For each woman compressed age,
body weight and height,
breast composition category according to ACR,
compressed breast thickness,
and exposure parameters (kVp,
mAs; as displayed onto the unit) for each of two standard views for each breast were recorded.
Air-kerma and average glandular dose were returned from the parameters,
and mean values and standard deviation were calculated.
By use of manual technique instead of AEC-device up to 45% of air-kerma and 51% reduction of AGD were achieved,
despite of the fact that mean compressed breast thickness in „manual“ group was 7% higher than in first group (Table 3 and Table 4).
We assume that lower MGDs in manual technique group was the result of at least two factors: firstly,
experienced RTs were able to estimate breast composition and structure by use of palpation and visual inspection,
more successfuly than younger colleagues; secondly,
their superior breast positioning skills enabled better breast evening onto the support table resulting in optimisation of exposure.