The Clinical Study Of Low Contrast-and Low Radiation Dose Of Coronary CT Angiography

Part oneThe feasibility of ultra-low iodine contrast material of coronary computed tomography angiographyPurpose:To investigate the feasibility of coronary computed tomography angiography (CCTA) using30ml of contrast material with a concentration of270mg I/ml and a tube voltage of80kVp combined with prospectively ECG-triggered, high pitch and iterative reconstruction compared to the standard CCTA protocol using60ml of contrast material with a concentration of370mg I/ml and a tube voltage of100kVp.Materials and Methods:A total of82patients (33women,49men; age range35-84years; mean age,56years) with body mass index (BMI) less than25kg/m2and heart rate (HR) lower than70beats per minute (bmp) were randomly assigned into two groups and underwent CCTA examination. Forty-three patients (21women,22men; age range36-77years; mean age,56years) were scanned using a tube voltage of100kVp after administration of60ml of a high concentration of iodine contrast material (Iopromide,370mg I/ml, Bayer, Berlin, Germany)(100kVp group).39patients (12women,27men; age range35-84years; mean age,55years) were scanned using a low tube voltage of80kVp after administration of30ml of a low concentration of iodine contrast material (iodixanol,270mg I/ml, GE Healthcare)(80kVp group). Images acquisition were performed under prospective ECG-triggered, high-pitch (3.4) mode. Automatic tube current modulation (ATCM, CARE Dose4D; Siemens Medical Solutions) was used in both acquisitions. The other scan parameters were as follows: tube rotation time,0.28seconds; collimation,0.6mm x64x2. contrast enhancement was achieved by injecting contrast material at5ml/s followed by a40ml saline flush at the same speed. The intensity of the injection-associated discomfort, defined as favor, coldness, pain in connection with the contrast medium injection, was asked and rated by patients on a three-point scale:none, mild and severe. CT images were reconstructed with filtered back projection (FBP) in100kVp group and with sinogram-affirmed iterative reconstruction (SAFIRE) in80kVp group. For the SAFIRE algorithm, five presets are available for adapting the noise model and controlling image impression and noise reduction. A medium strength level of3was used in our study. One observer measured intravascular attenuations and standard deviations (SD) within ascending aorta (AA), proximal right coronary artery (RCA), proximal left main coronary artery (LMA), proximal left anterior descending artery (LAD), proximal left circumflex artery (LCX) and adjacent perivascular tissue. The signal-noise-ratio (SNR, SNR=attenuation of the vessel lumen/lumen noise) and contrast-noise-ratio (CNR, CNR=(attenuation of the vessel lumen-attenuation of adjacent perivascular tissue)/image noise) were calculated. Two radiologists reviewed all images and scored them using a4-point scale:A score of1corresponded to a lack of vessel wall definition due to marked motion artifact, poor vessel opacification, prominent structural discontinuity, or high image noise rendering the segment non-diagnostic; a score of2corresponded to some motion artifacts or image noise, fair vessel opacification, or minimal structural discontinuity; a score of3corresponded to minor motion artifacts or image noise, good vessel opacification, and no structural discontinuity; and a score of4corresponded to the absence of motion artifacts or minimal noise, excellent vessel opacification, and no structural discontinuity. In the event of observer disagreement, another reading session was convoked to reach an agreement. A per-vessel and per-patient image quality score was defined as the worst score found in any segment for each vessel or each patient. The effect of HR and BMI on the image quality was further analyzed. The CT volume dose index (CTDIvol) and dose-length product (DLP) were recorded and the effective dose (ED) and size-specific dose estimate (SSDE) of each patient were calculated. The objective image quality (CT value, SNR, CNR) and radiation dose (CTDIvol, DLP, ED, SSDE) were compared with independent-sample t test and the subjective image quality was compared with independent samples nonparametric test. The frequency of discomfort was compared between the two contrast media groups using Fisher’s exact test.Results:The mean attenuation in80kVp group (365±72HU) was significantly lower than in100kVp group (507±69HU, P<0.001), as well as the image noise (22±5HU vs.25±4HU, P=0.003). SNR in80kVp was significantly lower than in100kVp group (17.4±4.7,20.8±3.7, P<0.001) while CNR showed no significant difference between the two groups (80kVp vs.100kVp group,22.6±6.2vs.24.5±4.0, P=0.099). A total of1014segments (538segments for100kVp group,476segments for80kVp group) with a diameter>1.5mm in82patients were assessed for subjective image quality, while298segments were missed due to anatomical variants or a vessel diameter of less than1.5mm. Among the538segments of100kVp group,15segments (3%) were scored1,17segments (3%) scored2,189segments (35%) scored3and317segments (59%) scored4. Among the476segments of80kVp group,16segments (3%) were scored1,26segments (6%) scored2,143segments (30%) scored3and291segments (61%) scored4. The inter-observer agreement was moderate for100kVp group (k=0.588, P<0.001) and good for80kVp group (k=0.620, P<0.001). On a per-patient basis, there was no difference for subjective image quality among the two groups (100kVp vs.80kVp group,2.5±0.9vs.2.4±0.9, P=0.764). On a per-coronary basis, the mean scores of100kVp group and80kVp were2.9±1.1and2.9±1.0for RCA (P=0.984),3.8±0.5and3.7±0.6for LMA (P=0.556),3.2±0.5and3.1±0.8for LAD (P=0.991),2.9±0.8and2.9±0.9for LCX (P=0.696), which showed no significant difference between the two groups. The CT value of LCX and image noise showed a significant difference between the patients with HR less than65bpm and patients with HR higher than65bpm in100kVp group, while no significant difference was found for other measurements in100kVp group. In80kVp group, no significant difference was found for all measurements (CT value, SNR and CNR) between the patients with HR less than65bpm and patients with HR higher than65bpm. In100kVp group, there was no significant difference in the subjective image quality between the patients with HR less than65bpm and patients with HR higher than65bpm. However, for the80kVp group, a higher score was found in RCA of patients with HR less than65bpm compared to the patients with HR higher than65bpm (3.1±0.9vs.2.1±1.1, P=0.014). For100kVp group and80kVp group, a lower image noise was found in patients with BMI less than23kg/m2compared to the patients with BMI higher than23kg/m2(100kVp group:24±4HU vs.27+5HU, P=0.01;80kVp group:20±4HU vs.25±3HU, P<0.001). For80kVp group, the patients with BMI less than23kg/m2showed higher SNR and CNR. No significant difference was found for subjective image quality between the patients with BMI less than23kg/m2and BMI higher than23kg/m2in the two groups. The frequency and intensity of injection-associated discomfort in80kVp group was lower than in100kVp group (P<0.001).The amount of iodine administrated and the radiation dose in80kVp group were reduced by64%and54%respectively when compared with100kVp group.Conclusions:Combined with prospectively ECG-triggered high-pitch technology and iterative reconstruction, CCTA using30ml of contrast material with a concentration of270mg I/ml is feasible for patients with BMI less than25kg/m2and HR lower than65bmp with substantial reduction in radiation dose and contrast material. Part twoThe feasibility of coronary CT angiography using a tube voltage of70kVp and30ml of contrast materialPurpose:To investigate the feasibility of CCTA using70kVp tube voltage and30ml contrast material combined with prospectively ECG-triggered, high-pitch and iterative reconstruction compared to the standard CCTA protocol using60ml of contrast material and a tube voltage of100kVp.Materials and Methods:A total of80patients (42men,38women, mean age,58years; age range,36-81years) with BMI less than25kg/m2and HR lower than70bmp were randomly assigned into two groups and underwent CCTA examination. Forty patients (21men,19women; age range,36-76years; mean age,56years) were scanned using a tube voltage of100kVp after administration of60ml of iodine contrast material (100kVp group). Another40patients (21men,19women; age range,43-81years; mean age,60years) were scanned using a low tube voltage of70kVp after administration of30ml of iodine contrast material (70kVp group). Images acquisition were performed under the mode of prospective ECG-triggered and high pitch (3.4). ATCM (CARE Dose4D) was used in both acquisitions. The other scan parameters were as follows: tube rotation time,0.28seconds; collimation,0.6mm x64x2. Contrast enhancement was achieved by injecting contrast material at5ml/s followed by a40ml saline flush at the same speed. CT images were reconstructed with FBP in100kVp group and and with SAFIRE in70kVp group with a medium strength level of3. The mean CT attenuations and standard deviations (SD) of coronaries (proximal RCA, proximal LMA, proximal LAD and proximal LCX) and adjacent perivascular tissue were measured and SNR and CNR were calculated. Two radiologists reviewed all images and scored them using a4-point scale (as above). The effect of HR and BMI on the image quality was further analyzed. The CTDIvol and DLP were recorded and the ED and SSDE of each patient were calculated. The objective image quality (CT value, SNR, CNR) and radiation dose (CTDIvol, DLP, SSDE) were compared with independent-sample t test and the subjective image quality was compared with independent samples nonparametric test.Results:The mean attenuation in70kVp group (603±86HU) was significantly higher than in100kVp group (503±68HU, P<0.001), as well as the image noise (42±5HU vs.25±4HU, p<0.001). SNR and CNR in70kVp (14.6±3.0,17.8±3.4) was significantly lower than in100kVp group (20.7±3.8,24.4±4.1, P<0.001). Among the80patients, a total of987coronary segments were assessed while213segments were missed due to anatomical variants or a vessel diameter of less than1.5mm. Among the499segments of100kVp group,15segments (3%) were scored1,17segments (4%) scored2,181segments (36%) scored3and286segments (57%) scored4. Among the488segments of70kVp group,5segments (1%) were scored1,34segments (7%) scored2,161segments (33%) scored3and288segments (59%) scored4. The inter-observer agreement was moderate for100kVp group (k=0.560, P<0.001) and good for70kVp group (k=0.639, P<0.001). On a per-patient basis, there was no difference for image quality between the two groups (100kVp group vs.70kVp group,2.42±0.93vs.2.42±0.68, P=0.644). On a per-coronary basis, the subjective image quality in100kVp group and70kVp group were2.9±1.1and3.2±0.7for RCA (P=0.416),3.8±0.5and3.9±0.3for LMA(P=0.509),3.2±0.5and3.2±0.6for LAD (P=0.923),2.9±0.7and2.7±0.8for LCX (P=0.102), which showed no significant difference between the two groups. No difference was found for all measurements (CT value, SNR and CNR) between the patients with HR less than65bpm and patients with HR higher than65bpm in two groups, as well as the subjective image quality in100kVp group. However, for the70kVp group, a higher score in RCA was found in patients with HR less than65bpm compared to the patients with HR higher than65bpm (3.3±0.7,2.8±0.7, P=0.027). For100kVp group and70kVp group, a lower image noise was found in patients with BMI less than23kg/m2compared to the patients with BMI higher than23kg/m2(100kVp group:24+3HU vs.27±5HU, P=0.021;70kVp group:41±5HU vs.44+4HU, P=0.04). The SNR of RCA in100kVp group and the CT value of RCA and CNR of RCA, LMA and LAD in70kVp group were significantly higher in patients with BMI less than23kg/m2compared to patients with BMI higher than23kg/m2. No significant difference was found for subjective image quality between the patients with BMI less than23kg/m2and patients with BMI higher than23kg/m2in two groups. The radiation dose and contrast material in patients of70kVp group were reduced by76%and50%respectively when compared with100kVp group.Conclusions:In conclusion, our study shows that prospectively ECG-triggered high-pitch CCTA at70kVp with30mL iodinated contrast medium volume can obtain diagnostic image quality with substantially reduced radiation dose in selected patients with BMI less than25kg/m2and HR less than65bpm.