Our Radiology department has three CT scanners models from different brands with angular and longitudinal tube current modulation,
no iterative reconstruction,
but distinct detector collimation: 64 channels,
16 channels and 8 channels.
For convenience we will name them as CT64,
CT16 and CT8,
respectively.
We collected acquisition parameters and dose information from randomly selected examinations,
totalling 510 CT exams (70 head exams,
50 chest exams and 50 abdomen exams for each CT scanner) scanned on 2018 accounted only for patients over 18 years and without contrast administration.
We chose these anatomical regions since they represent about 84% of the total number of exams in the radiology department.
The parameters collected were: patient weight,
patient effective diameter,
acquisition protocol,
peak voltage,
tube current,
rotation time,
pitch,
slice width.
To settle the adult standard patient we accessed the weight and measured the effective diameter on axial images.
The patient effective diameter was accessed only for chest and abdomen.
It was determined as the square root of the lateral dimension times the anteroposterior dimension.
chest measurements were done on carina of trachea level and abdomen measurements were at lumbar spine L4-L5 transition level.
The standard patient weight and effective diameter ranges were set as mean ± 1 standard deviation.
The CTDIvol per scan and DLP per scan was accessed at the picture and archiving information system (PACS) and we computed the median and 75th percentile for the following groups:
Head: Data from all head exams collected; Chest-W: Data from chest exams with patient’s weight within determined range; Chest-ED: Data from chest exams with patient’s effective diameter within determined range; Abdomen-W: Data from abdomen exams with patient’s weight within determined range; Abdomen-ED: Data from abdomen exams with patient’s effective diameter within determined range.
Adult standard patient weight range was set at 54-88 Kilograms and effective diameter range was set at 25 - 33 centimeters for both chest and abdomen.
Considering these ranges the total number of exams in each group was: Head (202); Chest-W (94); Chest-ED (126); Abdomen-W (104); Abdomen-ED (118).
Median and 75th percentile for CTDIvol and DLP computed for each group was:
|
CTDIvol (mGy)
|
DLP (mGy.cm)
|
Median
|
75th percentile
|
Median
|
75th percentile
|
Head
|
66,2
|
77,3
|
1173,875
|
1503,6
|
Chest-W
|
9,32
|
16,83
|
357,63
|
563
|
Chest-ED
|
9,44
|
16,1
|
353,68
|
553,16
|
Abdomen-W
|
13,63
|
21,7
|
641,23
|
824,76
|
Abdomen-ED
|
14,07
|
24,2
|
671,58
|
878,22
|
According to ICRP DRL values are set at the 75th percentile of collected data.
To evaluate the impact of choosing weigh against effective diameter,
we compared the difference in CTDIvol and DLP between groups.
|
CTDIvol (mGy)
|
DLP (mGy.cm)
|
Median
|
75th percentile
|
Median
|
75th percentile
|
Difference between Chest Groups
|
0,125 (1,34%)
|
0,725 (4,31%)
|
3,945 (1,10%)
|
9,84 (1,75%)
|
Difference between Abdomen Groups
|
0,44 (3,23%)
|
2,5 (11,52%)
|
30,355 (4,73%)
|
50,465 (6,12%)
|
Values of median and 75th percentile found for Chest-W and Chest-ED were similar,
as found for Abdomen-W and Abdomen-ED (Fig. 1 ; Fig. 2).
The biggest difference was found comparing the 75th percentile of the Abdomen groups,
in which the difference was 11,52%.
Even so all differences were relatively low,
especially in Chest groups (all below 5%).
Fig. 2 shows several outliers for DLP on Abdomen groups.
To determine the cause of these outlier each equipment was analyzed individually for Head,
Chest-W and Abdomen-W groups.
(Fig. 3 ; Fig. 4 ; Fig. 5).
As it can be seen on Fig. 3 ,
Fig. 4 and Fig. 5 CTDIvol varied significantly for each equipment.
CT64 showed higher CTDIvol then the other scanners in all groups.
On Head group data indicates that CT8 uses dose modulator while CT16 and CT64 don’t.
Data indicates,
in general,
that there is a lack of standardization on equipment protocols.