Radiation dose reduction in CTCA
In a single centre cohort study we demonstrated that the combination of iterative reconstruction and automated tube current selection could reduce radiation dose while maintaining image quality.
(Williams et al.,
2013b)
We reviewed the images of 942 consecutive patients who underwent electrocardiogram-gated CTCA using a 320 multidetector scanner (Aquilion ONE,
Toshiba Medical Systems).
The first 228 patients (Group 1) had their images reconstructed using filtered back projection (QDS+).
The next 379 patients had their images reconstructed using iterative reconstruction (AIDR,
Adaptive Iterative Dose Reduction).
The final 335 patients had their images reconstructed using a new iterative reconstruction (AIDR3D,
Adaptive Iterative Dose Reduction Three dimensional).
Tube current was selected based on BMI for Groups 1 and 2 and selected automatically based on scout image attenuation for Group 3 (Table 1).
Subjective image quality was graded on a four-point scale (1 excellent,
4 non-diagnostic).
There were no differences in age (P=0.975),
body mass index (P=0.435) or heart rate (P=0.746) between groups.
Image quality improved with iterative reconstruction and automatic tube current selection (1.3 (95% confidence intervals: 1.2,
1.4),
1.2 (1.1,
1.2) and 1.1 (1.0,
1.2) respectively; P <0.001) and radiation dose decreased (274 (260,
290),
242 (230,
253) and 168 (156,
180) mGy.cm respectively,
P <0.001) (Figure 2).
Example images can be seen in Figure 3.
Therefor the application of these techniques in CTCA led to a 39% reduction in radiation dose while maintaining image quality.
Table 1: Techniques applied in each group
|
Group 1 |
Group 2 |
Group 3 |
Reconstruction |
Filtered back projection |
AIDR iterative reconstruction |
AIDR3D iterative reconstruction |
Tube current selection |
Based on body mass index |
Based on body mass index |
Automated selection of tube current based on scout image attenuation |
Radiation dose reduction in CT myocardial perfusion imaging
The techniques that are used to minimize radiation dose in CTCA can also be applied to CT myocardial perfusion imaging.
In a single centre cohort study we assessed the application of iterative reconstruction and automated tube current selection on radiation dose and image quality in CT myocardial perfusion imaging.
(Williams,
Golay,
et al.,
2013a)
We reviewed the images of 56 patients who underwent “snap shot” rest and adenosine stress CT myocardial perfusion imaging using a 320-multidetector scanner (Aquilion ONE,
Toshiba Medical Systems).
The first 28 patients (Group 1) underwent imaging with a standard protocol and for the second 28 patients (Group 2) an optimized protocol was used (Table 2).
Dose length product (DLP) and image noise were recorded.
Subjective image quality was assessed (from 1 (excellent) to 4 (uninterpretable)) and summed for the 17 myocardial segments.
There was no difference in gender,
body mass index,
heart rate at rest or stress,
or z-axis collimation at rest or stress between patients in Group 1 and Group 2.
There was a reduction in tube current (467 (449,485) versus 318 (245,392) mA,
P<0.001),
and tube voltage (100% versus 33% at 120 kV,
P <0.001) in Group 2.
DLP was lower in Group 2 during rest (224 (210,238) versus 141 (109,173) mGy.cm,
P <0.001) and stress imaging (604 (486,722) versus 232 (177,286) mGy.cm,
P <0.001,
Figure 4).
There was no difference in subjective image quality between groups.
Therefore the application of these techniques led to a 60% reduction in radiation dose in CT myocardial perfusion imaging,
while maintaining image quality.
Table 2: Techniques applied in each group.
|
Group 1 |
Group 2 |
Reconstruction algorithm |
Filtered back projection (QDS+) |
Iterative reconstruction (AIDR3D) |
Tube voltage |
120kV for all |
Tailored to body mass index |
Segments for reconstruction |
Multisegment if heart rate >65 beats per minute |
Half segment reconstruction |
Dose modulation |
None |
For stress imaging with 70-80% of the RR interval at full dose and the rest of the cardiac cycle at low dose |
Iterative reconstruction and CACS
The techniques used to minimize radiation dose in other forms of cardiovascular CT can be applied to CACS,
but their potential to influence calcium quantification needs to be considered. In a single centre cohort study we assessed the feasibility of CACS radiation dose reduction using lower tube current and iterative reconstruction.
(Rodrigues et al.,
2014)
Artificial noise was added to raw data from 27 CACS studies from symptomatic patients to simulate lower tube current scanning (75%,
50% and 25% original current).
All studies were performed on a 320-multidetector scanner (Aquilion ONE,
Toshiba Medical Systems) at 120kVp.
Data was reconstructed using filtered back projection (QDS+) and iterative reconstruction (AIDR3D with mild,
standard and strong levels).
Agatston scores were independently measured by two readers and CACS percentile risk scores were calculated.
At 75%,
50% and 25% tube currents all AIDR3D reconstructions decreased image noise relative to QDS+ (P<0.05) (Figure 5).
All AIDR3D reconstructions resulted in small reductions in Agatston score relative to QDS+ at standard tube current (P<0.05).
Agatston scores increased with QDS+ at 75%,
50% and 25% tube current (P<0.05),
whereas no significant change was observed with AIDR3D-mild at any tested tube current.
No difference in percentile risk score with AIDR3D-mild at any tube current occurred compared with QDS+ at standard tube current (P>0.05).
Inter-observer agreement for AIDR3D-mild remained excellent even at 25% tube current (Intraclass Correlation Coeffiecient 0.997).
Therefore AIDR3D-mild permits up to 75% reduction in CACS tube current whilst maintaining excellent intra- and inter-observer variability and without altering risk classification.