Compliance to suggested table
In phase I,
949 patients were examined: 650 were “Small”, 277 “Medium” and only 22 “Large”.
248 examinations were performed with the “Small Protocol” (100 kVp),
641 with the “Medium Protocol” and 60 with the “Large Protocol”.
As shown in table 2,
481 out of 949 acquisitions (50.7%) were not optimized,
i.e.
proper protocols were not chosen according to patient BMI,
as proposed in the suggested chart (Figure 1).
In particular,
412 Small patients out of 650 (63.4%) were imaged with 120 kVp instead of 100 kVp.
The situation was much better for “Medium” and “Large” patients (80% and 68% acquired correctly,
respectively).
Visual representation of phase I data is shown in Figures 2 and 3.
Acquisitions |
BMI<26 |
26<BMI36 |
BMI>36 |
Total |
Protocol Small - 100 kVp |
231 |
17 |
0 |
248 |
Protocol Medium - 120 kVp |
412 |
222 |
7 |
641 |
Protocol Large - 140 kVp |
7 |
38 |
15 |
60 |
Total |
650 |
277 |
22 |
|
Table 2 – Acquisitions data for phase I.
In phase II,
451 patients were examined: 287 “Small”,
151 “Medium” and 13 “Large”.
202 examinations were performed with the “Small Protocol” (100 kVp),
226 with the “Medium Protocol” and 23 with the “Large Protocol”.
Table 3 shows that 125 out of 451 acquisitions (28%) were not performed correctly.
Considering “Small” patients,
100 out of 287 (35%) were imaged with 120 kVp instead of 100 kVp.
83% of “Medium” patients (126/151) were acquired correctly,
and 100% of “Large” patients were properly examined with 140 kVp.
Acquisitions |
BMI<26 |
26<BMI36 |
BMI>36 |
Total |
Protocol Small - 100 kVp |
187 |
15 |
0 |
202 |
Protocol Medium - 120 kVp |
100 |
126 |
0 |
226 |
Protocol Large - 140 kVp |
0 |
10 |
13 |
23 |
Total |
287 |
151 |
13 |
|
Table 3 – Acquisitions data for phase II.
Exposure analysis
Exposure analysis was carried out for all phases,
comparing two subgroups: “Small” patients acquired with 100 kVp (Small-100,
hereafter) and “Small” patients acquired with 120 kVp (Small-120,
hereafter).
Acquisition phase
|
Small 100 kVp
CTDIvol (mGy)
|
Small 120 kVp
CTDIvol (mGy)
|
Unenhanced phase
|
255
|
263
|
Arterial phase
|
322
|
365
|
Portal phase
|
418
|
512
|
Table 3 - Number of acquisitions collected for the analysis.
Acquisition phase
|
Small 100 kVp
CTDIvol (mGy)
|
Small 120 kVp
CTDIvol (mGy)
|
Unenhanced phase
|
6.4 (6.0 – 7.6)
|
9.5 (9.2 – 10.9)
|
Arterial phase
|
6.3 (6.0 – 7.4)
|
9.5 (9.1 – 10.6)
|
Portal phase
|
7.9 (7.0 – 9.2)
|
10.7 (9.9 – 12.0)
|
Table 4 - Median and IQR for CTDIvol (mGy) of two subgroups.
Table 4 shows median and interquartile range of CTDIvol for different group and phases.
For patient Small scanned with 120 kVp,
CTDIvol is up to 48% higher,
differences are statistically significant for all phases (p < 0.01).
Differences in CTDIvol are obviously reflected on DLP and effective dose (see Tables 5 and 6).
In particular,
for 120 kVp protocols,
DLP increase is up to 51% and effective dose increase is up to 48%.
Acquisition phase |
Small 100 kVp
CTDIvol (mGy) |
Small 120 kVp
CTDIvol (mGy) |
Unenhanced phase |
186.4 (167.1 – 227.7) |
282.0 (261.0 – 331.6) |
Arterial phase |
179.8 (161.3 – 216.8) |
276.8 (256.0 – 315.1) |
Portal phase |
535.5 (467.3 – 640.5) |
746.7 (679.5 – 857.0) |
Table 5 - Median and IQR for DLP (mGy cm) of two subgroups.
Acquisition phase
|
Small 100 kVp
CTDIvol (mGy)
|
Small 120 kVp
CTDIvol (mGy)
|
Unenhanced phase
|
4.1 (3.7 – 4.7)
|
6.1 (5.6 – 6.8)
|
Arterial phase
|
4.0 (3.6 – 4.5)
|
6.0 (5.4 – 6.6)
|
Portal phase
|
9.8 (8.9 – 10.8)
|
13.6 (12.1 – 14.9)
|
Table 6 - Median and IQR for E (mSv) of two subgroups.
Visual representation of CTDIvol and effective dose in the 3 phases for the 2 subgroups are shown in figg 4 and 5.
Image quality
All images were evaluated as diagnostic (score always ≥ 2).
Mean scores for the 2 readers are shown in the following tables.
Both readers evaluated that soft tissue visualization was better with 120 kVp rather than 100 kVp (p < 0.05).
In addition,
for reader 1,
lung parenchyma was imaged better with 120 kVp than with 100 kVp and the difference,
despite being small,
was found to be statistically significant.
The other differences,
especially those referring to the overall image quality,
were not statistically significant.
The agreement of the 2 readers was slight (0.14),
due to the lack of reference images.
Reader 1
(10y of experience)
|
Lung parenchyma
|
Soft Tissue
|
Bone visualization
|
Overall Score
|
Small
100 kVp
|
4.8*
|
3.7*
|
4.5
|
4.3
|
Small
120 kVp
|
5.0*
|
4.1*
|
4.5
|
4.5
|
Table 7 - Mean scores for reader 1.
*indicates statistical difference between the two results.
Reader 2
(1y of experience)
|
Lung parenchyma
|
Soft Tissue
|
Bone visualization
|
Overall Score
|
Small
100 kVp
|
3.7
|
3.5*
|
3.7
|
3.7
|
Small
100 kVp
|
3.7
|
3.8*
|
3.8
|
3.9
|
Table 8 - Mean scores for reader 2.
*indicates statistical difference between the two results.