The present work relies on dosimetric approach validated in a previous publication [13],
where a method enabling the direct measurement of CTDIvol and DLP for helical CT scans has been reported.
This methodology involves the use of a measuring system suspended at the CT isocenter and uncoupled to the couch displacement.
The measuring system does not shift with the table,
but is suspended at the CT isocenter and free from table movement.
During the scan,
the beam is oriented towards the ionization chamber center for the entire scan time.
In this way,
also in the case of helical scans,
all dose contributions (including overscanning) are integrated on the ion chamber length,
as for CTDI for sequential scans,
and the dose of the whole scan can be acquired.
With reference to Figure 1,
the exposed scan time tSCAN is not equal to the imaged scan time tIM (the time required to scan the length of the object to be imaged),
because each section requires the same number of projection for reconstruction.
The exposed scan time consists of the imaged scan time tIM and of the overscan time tOV,
so that tSCAN= tIM+tOV.
In the following,
the symbol OV will be used to indicate the CT overscanning.
As demonstrated in the previous paper [4],
is assessed by multiplying the ion chamber length L by the dose rate integral in a time interval of duration (grey areas of Figure 1).
DLPOV is then expressed as a weighted average of the values achieved at center and periphery.
A dual source SOMATOM Definition Flash multislice helical CT (Siemens Healthcare,
Germany) has been used in the current study.
The DLPOV has been evaluated by means of Food and Drug Administration head and body phantoms [14] and of a pencil chamber Radcal 10X6-3CT with a charge collection length .
The measuring system,
composed by the phantom and the pencil chamber,
has been suspended at the gantry isocenter in a stationary longitudinal position.
All scans have been performed in helical mode with a beam collimation of for all scans.
Rotation times of 0.28 s,
0.33 s,
0.5 s and 1 s have been investigated.
As resumed in Tables 1-2,
a wide range of pitch values has been used depending on the selected acquisition protocol.
Figure 2 shows the overscan length as a function of pitch for different rotation times (head phantom).
The overscan length has shown a positive correlation for pitch values ranging from 0.35 to 1.0; it becomes slightly constant for pitch values higher than 1.0.
Figure 3 shows the overscan length as a function of X-ray tube rotation time from 0.28 to 1 s per rotation,
for different pitches.
Overscan length has remained constant with X-ray tube rotation times,
even at different pitches,
with measurements varying only 7% for lowest pitches.
Moreover,
the contribution of decreases from 36% to 4% increasing the overscan length from 2.5 to 30 cm (Figure 4).
All DLP values are affected by ±2% of error associated to the ion chamber readings.
Figure 5 illustrates the dependence of DLPOV,
normalized to 100 mAseff,
from acquisition parameters.
The dependence of absolute DLPOV value from pitch was obtained for different rotation times and acquisition protocols. At a fixed length of reconstructable volume,
the absolute value of the dose due to overscanning increases as the pitch values is increased.