64-slice multidetector coronary CT angiography: in vitro evaluation of 68 different stents

The purpose of this study was to test a large sample of different coronary artery stents using four image reconstruction approaches with respect to lumen visualization, lumen attenuation, and image noise in 64-slice multidetector-row computed tomography (MDCT) in vitro and to provide a catalogue of currently used coronary artery stents when imaged with state-of the-art MDCT. We examined 68 different coronary artery stents (57 stainless steel, four cobalt-chromium, one cobalt-alloy, two nitinol, four tantalum) in a coronary artery phantom (vessel diameter 3 mm, intravascular attenuation 250 HU, extravascular density −70). Stents were imaged in axial orientation with standard parameters: 32x0.6 collimation, pitch 0.24, 680 mAs, 120 kV, rotation time 0.37 s. Four different image reconstructions were obtained with varying convolution kernels and section thicknesses: (1) soft, 0.6 mm, (2) soft, 0.75, (3) medium soft, 0.6, and (4) stent-optimized sharp, 0.6. To evaluate visualization characteristics of of the stent, the lumen diameter, intraluminal density and noise were measured. The high-resolution kernel offered significantly better average lumen visualization (57% ±10%) and more realistic lumen attenuation (222 HU ±66 HU) at the expense of increased noise (15.3 HU ±3.7 HU) compared with the soft and medium-soft CT angiography (CTA) protocol (p<0.001 for all). Stents with a lumen visibility of more than 66% were: Arthos pico, Driver, Flex, Nexus2, S7, Tenax complete, Vision (all 67%), Symbiot, Teneo (70%), and Radius (73%). Only ten stents showed a lumen visibility of less than 50%. Stent lumen visibility largely varies depending on the stent type. Even with the improved spatial resolution of 64-slice CT, a stent-optimized kernel remains beneficial for stent visualization when compared with the standard medium-soft CTA protocol. Using 64-slice CT and high-resolution kernel, the majority of stent products show a lumen visibility of more than 50% of the stent diameter.     

Improved coronary artery stent visualization and in-stent stenosis detection using 16-slice computed-tomography and dedicated image reconstruction technique

OBJECTIVE: The aim of this study was to compare the visualization of different coronary artery stents and the detectability of in-stent stenoses during 4-slice and 16-slice computed tomography (CT) angiography in a vessel phantom. MATERIAL AND METHODS: Ten coronary stents were introduced in a coronary artery vessel phantom (plastic tubes with an inner diameter of 3 mm, filled with iodinated contrast material diluted to 220 Hounsfiled Units [HU], surrounded by oil [60 HU]). CT scans were obtained perpendicular to the stent axes on a 4-slice scanner (detector collimation 4x1 mm; table feed 1.5 mm/rotation, mAs 300, kV 120, medium-smooth kernel) and a 16-slice scanner (detector collimation 12x0.75 mm; table feed 2.8 mm/rotation, mAs 370, kV 120, reconstruction with a standard and an optimized sharp kernel). Longitudinal multiplanar reformations were evaluated regarding visible lumen diameters and intraluminal attenuation values. Additionally, the stents were scanned with the same parameters after implantation of 60% stenoses (HU 30). RESULTS: Using the same medium-smooth kernel reconstruction with 4-slice and 16-slice CT, there was a slight increase in the average visible lumen area (26% versus 31%) and less increase of average intraluminal attenuation values (380 HU versus 349 HU). Significant improvement of lumen visualization (54%, P<0.01) and attenuation values (250, P<0.01) was observed for the 16-slice scans using the sharp kernel reconstruction. In-stent stenoses could be more reliably identified (or ruled out) by 16-slice CT and sharp reconstruction kernel when compared with the other 2 methods. CONCLUSION: 16-slice CT using a dedicated sharp kernel for image reconstruction facilitates improved visualization of coronary artery stent lumen and detection of in-stent stenoses.     

Update on multidetector coronary CT angiography of coronary stents: in vitro evaluation of 29 different stent types with dual-source CT

The aim of this study was to test a large sample of the latest coronary artery stents using four image reconstruction approaches with respect to lumen visualization, lumen attenuation, and image noise in dual-source multidetector row CT (DSCT) in vitro and to provide a CT catalogue of currently used coronary artery stents. Twenty-nine different coronary artery stents (19 steel, 6 cobalt-chromium, 2 tantalum, 1 iron, 1 magnesium) were examined in a coronary artery phantom (vessel diameter 3 mm, intravascular attenuation 250 HU, extravascular density −70 HU). Stents were imaged in axial orientation with standard parameters: 32 × 0.6 collimation, pitch 0.24, 400 mAs, 120 kV, rotation time 0.33 s. Image reconstructions were obtained with four different convolution kernels (soft, medium-soft, standard high-resolution, stent-dedicated). To evaluate visualization characteristics of the stent, the lumen diameter, intraluminal density, and noise were measured. The stent-dedicated kernel offered best average lumen visualization (54 ± 8.3%) and most realistic lumen attenuation (222  ± 44 HU) at the expense of increased noise (23.9 ± 1.9 HU) compared with standard CTA protocols (p < 0.001 for all). The magnesium stent showed the least artifacts with a lumen visibility of 90%. The majority of stents (79%) exhibited a lumen visibility of 50–59%. Less than half of the stent lumen was visible in only six stents. Stent lumen visibility largely varies depending on the stent type. Magnesium is by far more favorable a stent material with regard to CT imaging when compared with the more common materials steel, cobalt-chromium, or tantalum. The magnesium stent exhibits a lumen visibility of 90%, whereas the majority of the other stents exhibit a lumen visibility of 50–59%. David Maintz and Matthias Burg contributed equally to this publication.     

Optimal scanning protocols of 64-slice CT angiography in coronary artery stents: An in vitro phantom study

Purpose

The purpose of the study was to investigate the optimal scanning protocol of 64-slice CT angiography for assessment of coronary artery stents based on a phantom study.

Materials and methods

Coronary stents with a diameter of 2.5 mm was implanted in thin plastic tubes with an inner diameter of 3.0 mm to simulate a coronary artery. The tubes were filled with iodinated contrast medium diluted to 178 HU, closed at both ends and positioned in a plastic container filled with vegetable oil (−70 to −100 HU). A series of scans were performed with a 64-slice CT scanner with the following protocols: section thickness: 0.67 mm, 1.0 mm, 1.5 mm, 2.0 mm, pitch value: 0.2, 0.3, 0.5 and reconstruction interval of 50% overlap of the section thickness. 2D axial and multiplanar reformatted images were generated to assess the visibility of stent lumen, while virtual intravascular endoscopy (VIE) was reconstructed to evaluate the artery wall and stent surface.

Results

Our results showed that a scanning protocol of 1.0 mm slice thickness with a pitch of 0.3 produced acceptable images with best demonstration of the intrastent lumen and stent surface with minimal image noise or artifacts. In contrast, submillimeter scans with 0.67 mm resulted in moderate artifacts which affected visualization of the coronary lumen, in addition to the increased noise. When the section thickness increased to 1.5 mm and 2.0 mm, visualization of the artery wall and stent surface was compromised, although the intrastent lumen was still visible.

Conclusion

Our in vitro study suggested that a scanning protocol of 1.0 mm section thickness with pitch of 0.3 is the optimal protocol for evaluation of coronary artery stents as it allows generation of acceptable images with better visualization of stent lumen, stent surface and coronary artery wall.     

CT angiography of various superficial femoral artery stents: an in vitro phantom study

Purpose

To evaluate CT reconstruction parameters to improve stent lumen visualization in vitro.

Material and methods

12 latest superficial femoral artery (SFA) stents were placed in a vessel phantom (diameter 4.7 mm, intravascular attenuation 250 HU, extravascular density 50 HU). Stents were imaged with a 128-slice scanner (SOMATOM Definition Flash, Siemens, Germany) with standard parameters: 120 kV, 200 mAs, collimation 128 mm × 0.6 mm. Different reconstruction parameters were evaluated: B26f, B30f, B45f, B46f and B60f kernel; slice thickness of 0.6, 2.0 and 5.0 mm. To measure visualization characteristics, stent lumen diameter and intraluminal attenuation were assessed.

Results

Best stent lumen visualization could be obtained using the B46f kernel (p < 0.001). The visible stent lumen ranged from 66.4% to 83.3% with a mean diameter of 77.7 ± 4.6%. Nitinol stents showed a significant improved lumen visibility compared to the cobalt–chromium stent (p = 0.02). The most realistic lumen attenuation was achieved using the B46f kernel with a mean attenuation of 259.3 ± 8.9 HU. The visible lumen diameter in protocols with 5 mm slice thickness was significantly lower (70.0 ± 4.9%) compared to thinner slices (p < 0.001).

Conclusion

CTA of SFA stents should be reconstructed with a slice thickness of 2.0 mm and a B46f kernel to achieve best image quality and to become more sensitive to exclude instent restenosis.     

Assessment of coronary artery stents using 16-slice MDCT angiography: evaluation of a dedicated reconstruction kernel and a noise reduction filter

To compare the effect of different reconstruction kernels and a noise-reducing postprocessing filter on the delineation of coronary artery stents in 16-slice CT-angiography. Ten patients with coronary stents (seven LAD, five RCX and three RCA) were examined with a 16-slice MDCT using standard acquisition parameters. Images were reconstructed using a medium soft (B30f) and a dedicated, edge-enhancing kernel (B46f). Additional postprocessing with an edge-preserving filter was performed on B46f images to reduce the image noise. In multiplanar reformations (MPRs) along and perpendicular to the stent axis, intraluminal attenuation values and the visible lumen diameter were measured. Image noise was measured in the subcutaneous fat using a region of interest (ROI) technique. Arterial enhancement in the aorta was 275.1 HU. Attenuation in the stent lumen was 390.4, 340.0 and 346.8 HU in MPRs derived from B30f, original B46 and postprocessed B46f images. The mean noise level was 20.4, 35.0 and 24.9 HU respectively. The visible lumen diameter was significantly greater in B46f and postprocessed B46f images (2.17 and 2.16 mm), compared to 1.93 mm in B30f images (p<0.01). Edge-enhancing reconstruction kernels increase the visible stent lumen, but also increase image noise. Dedicated postprocessing filters can reduce the introduced noise without a loss of spatial resolution.     

Coronary artery stent imaging with CT using an integrated electronics detector and iterative reconstructions: First in vitro experience

BackgroundDespite continuous improvements in CT technology, accurate stent lumen delineation remains challenging.PurposeThe aim was to evaluate the quality of coronary stent lumen delineation with CT using a detector with integrated electronics.MethodsTwelve coronary stents placed in plastic tubes and filled with contrast agent (CT number 250 HU) were imaged with either a 128-section dual-source CT machine equipped with conventional detector or with integrated electronics. On both scanners, images were reconstructed with filtered back projection (slice thickness 0.6 mm; increment 0.4 mm) and sinogram-affirmed iterative reconstruction (slice thickness 0.6 mm; increment 0.4 mm), and with iterative reconstruction (slice thickness 0.5 mm; increment 0.3 mm) on the integrated scanner. Two blinded, independent readers assessed image quality, noise, in-stent diameter, in-stent attenuation, and image sharpness by using signal intensity profiles across stents.ResultsInterreader agreement for image quality assessment was substantial (κ = 0.798). Both readers rated best image quality in data sets from integrated detector at highest spatial resolution (86 or 72% of stents rated best quality). Image noise was significantly lower in data sets scanned with integrated detector, being lowest at 0.6 mm slice thickness (14.3 vs 21.0 HU; P < .001). Differences between measured and true in-stent diameters and differences in attenuation across stents were smallest, and average/maximum image sharpness was highest in data sets from the integrated detector using iterative reconstructions.ConclusionCT coronary stent imaging is significantly improved by using a detector with integrated electronics combined with iterative reconstructions.     

Multislice CT angiography of the iliac arteries in the presence of various stents: in vitro evaluation of artifacts and lumen visibility.

RATIONALE AND OBJECTIVES: To evaluate the imaging characteristics of various iliac artery stents and stent-grafts in a multislice, computed tomography angiography (MSCTA) phantom study. METHODS: Twelve types of stents and three stent-grafts made of different materials (steel, nitinol, tantalum, cobalt-based alloy) were implanted in plastic tubes with an 8-mm inner diameter filled with iodinated contrast medium diluted to 200 HU. To evaluate the influence of scan parameters on artifacts, the Palmaz stent (as one example) was scanned with a four-slice scanner (Siemens VolumeZoom) with different detector collimations and pitches. All other stents were examined with a collimation of 4 x 1 mm and a table feed of 4 mm/rotation. Axial images and multiplanar reformations were evaluated regarding artifact size, lumen visibility, and intraluminal attenuation values. RESULTS: Higher pitch values caused more pronounced artifacts. Image quality and delineation of the stent struts improved with thinner detector collimation. The size of the stent-related artifacts and the visibility of the stent lumen depended on the underlying stent material and design. Pronounced artifacts, resulting in an insufficient delineation of the stent lumen, were caused by the Strecker tantalum stent and the Zenith stent. Moderate artifacts were caused by the Wallgraft, Passager, Palmaz P 395 and P 424, Bridge, Perflex, and ZA stents. Some artificial lumen narrowing but good lumen delineation was seen with the Strecker nitinol, Symphony, Memotherm, SMART, Corinthian, and Wallstent stents. CONCLUSIONS: Scanning parameters, stent material, and stent design influence lumen assessability and image quality in MSCTA. Detailed lumen assessment is impossible in the Strecker tantalum and Zenith stents and limited in the Wallgraft, Passager, Palmaz P 395 and PS 424, Bridge, Perflex, and ZA-stent stent-grafts but feasible in all other examined products.     

In vitro imaging of coronary artery stents: Are there differences between 16- and 64-slice CT scanners?

Purpose

To compare the performance of 64-slice with 16-slice CT scanners for the in vitro evaluation of coronary artery stents.

Methods and materials

Twelve different coronary artery stents were placed in the drillings of a combined heart and chest phantom, which was scanned with a 16- and 64-slice CT scanner. Coronal reformations were evaluated for artificial lumen narrowing, intraluminal attenuation values, and false widening of the outer stent diameter as an indicator of artifacts outside the stent.

Results

Mean artificial lumen narrowing was not significantly different between the 16- and 64-slice CT scanner (44% versus 39%; p = 0.408). The differences between the Hounsfield Units (HU) measurements inside and outside the stents were significantly lower (p = 0.001) with 64- compared to 16-slice CT. The standard deviation of the HU measurements inside the stents was significantly (p = 0.002) lower with 64- than with 16-slice CT. Artifacts outside the stents were not significantly different between the scanners (p = 0.866).

Conclusion

Visualization of the in-stent lumen is improved with 64-slice CT when compared with 16-slice CT as quantified by significantly lesser intraluminal image noise and less artificial rise in intraluminal HU measurement, which is the most important parameter for the evaluation of stent patency in vivo.     

High-resolution ex vivo imaging of coronary artery stents using 64-slice computed tomography—initial experience

The aim of the study was to evaluate the potential of new-generation multi-slice computed tomography (CT) scanner technology for the delineation of coronary artery stents in an ex vivo setting. Nine stents of various diameters (seven stents 3 mm, two stents 2.5 mm) were implanted into the coronary arteries of ex vivo porcine hearts and filled with a mixture of an iodine-containing contrast agent. Specimens were scanned with a 16-slice CT (16SCT) machine; (Somatom Sensation 16, Siemens Medical Solutions), slice thickness 0.75 mm, and a 64-slice CT (64SCT, Somatom Sensation 64), slice-thickness 0.6 mm. Stent diameters as well as contrast densities were measured, on both the 16SCT and 64SCT images. No significant differences of CT densities were observed between the 16SCT and 64SCT images outside the stent lumen: 265±25HU and 254±16HU (P=0.33), respectively. CT densities derived from the 64SCT images and 16SCT images within the stent lumen were 367±36HU versus 402±28HU, P<0.05, respectively. Inner and outer stent diameters as measured from 16SCT and 64SCT images were 2.68±0.08 mm versus 2.81±0.07 mm and 3.29±0.06 mm versus 3.18±0.07 mm (P<0.05), respectively. The new 64SCT scanner proved to be superior in the ex vivo assessment of coronary artery stents to the conventional 16SCT machine. Increased spatial resolution allows for improved assessment of the coronary artery stent lumen.     

Peripheral artery stent visualization and in-stent stenosis analysis in 16-row computed tomography: an in-vitro evaluation

The accuracy of 16-row multidetector CT in the visualization of different peripheral artery stents and in the appraisal of in-stent stenosis was assessed. Nine different stent types (nitinol and stainless steel) with three diameters (6, 8 and 10 mm) were used; altogether 27 stents were analyzed in a barrel-shaped vascular model. Low-grade (<40%) and high-grade (>60%) in-stent stenoses were simulated by polyurethane sticks (70 HU) of differing diameters (2–6 mm). Imaging was performed with 16×0.75-mm detector collimation, 130 mAs, 120 kV, 12-mm table feed/rotation, 1.0-mm slice thickness and 0.5-mm increment. The stent diameter, strut thickness, in-stent attenuation values, degree and degree of in-stent stenosis were evaluated. Nitinol stents showed significantly (P<10–6) less stent lumen narrowing, artificial strut thickening and overestimation of the degree of in-stent stenoses than stainless steel stents. In-stent attenuation values and artificial strut thickening were significantly (P<10–6) lower in 10- and 8-mm stents than in 6-mm stents. Stent lumen narrowing was significantly less in 10-mm stents than in 8-mm (P<10–4) or 6-mm (P<10–6) stents. In-stent stenoses were significantly overestimated, irrespective of the stent diameter. In 6-mm stents overestimation was significantly higher than in 8-mm (P<0.01) or 10-mm stents (P<10–6). Under in-vitro conditions 16-row MDCT allowed an accurate identification of in-stent stenosis, but significantly overestimated the effective degree of the stenosis.     

Coronary Artery Stent Evaluation Using a Vascular Model at 64-Detector Row CT: Comparison between Prospective and Retrospective ECG-Gated Axial Scans

Objective

We wanted to evaluate the performance of prospective electrocardiogram (ECG)-gated axial scans for assessing coronary stents as compared with retrospective ECG-gated helical scans.

Materials and Methods

As for a vascular model of the coronary artery, a tube of approximately 2.5-mm inner diameter was adopted and as for stents, three (Bx-Velocity, Express2, and Micro Driver) different kinds of stents were inserted into the tube. Both patent and stenotic models of coronary artery were made by instillating different attenuation (396 vs. 79 Hounsfield unit [HU]) of contrast medium within the tube in tube model. The models were scanned with two types of scan methods with a simulated ECG of 60 beats per minute and using display field of views (FOVs) of 9 and 18 cm. We evaluated the in-stent stenosis visually, and we measured the attenuation values and the diameter of the patent stent lumen.

Results

The visualization of the stent lumen of the vascular models was improved with using the prospective ECG-gated axial scans and a 9-cm FOV. The inner diameters of the vascular models were underestimated with mean measurement errors of -1.10 to -1.36 mm. The measurement errors were smaller with using the prospective ECG-gated axial scans (Bx-Velocity and Express2, p < 0.0001; Micro Driver, p = 0.0004) and a 9-cm FOV (all stents: p < 0.0001), as compared with the other conditions, respectively. The luminal attenuation value was overestimated in each condition. For the luminal attenuation measurement, the use of prospective ECG-gated axial scans provided less measurement error compared with the retrospective ECG-gated helical scans (all stents: p < 0.0001), and the use of a 9-cm FOV tended to decrease the measurement error.

Conclusion

The visualization of coronary stents is improved by the use of prospective ECG-gated axial scans and using a small FOV with reduced blooming artifacts and increased spatial resolution.     

Coronary artery stents in multislice computed tomography: in vitro artifact evaluation

RATIONALE AND OBJECTIVE: The aim of this study was to systematically compare the ability to assess the coronary artery lumen in the presence of coronary artery stents in multislice spiral CT (MSCT). METHODS: Ten different coronary artery stents were examined with 4- and 16-detector row MSCT scanners. For image reconstruction, a standard and a dedicated convolution kernel for coronary artery stent visualization were used. Images were analyzed regarding lumen visibility, intraluminal attenuation, and artifacts outside the stent lumen. Results were compared using repeated-measure analysis of variance. RESULTS: Depending on stent type, scanner hardware, and convolution kernel, artificial lumen narrowing ranged from 20% to 100%. The convolution kernel had the most significant influence on the visibility of the stent lumen. Artificial lumen narrowing and intraluminal attenuation changes decreased significantly using the dedicated convolution kernel. In general, most severe artifacts were caused by gold or gold-coated stents. CONCLUSIONS: Independent of the scanner hardware or dedicated convolution kernels, routine evaluation of most coronary artery stents is not yet feasible using MSCT.     

Revealing in-stent stenoses of the iliac arteries: comparison of multidetector CT with MR angiography and digital radiographic angiography in a Phantom model

OBJECTIVE: Our objective was to evaluate the detectability of in-stent stenoses in iliac artery stents using multidetector CT angiography in comparison with MR angiography and digital radiographic angiography. MATERIALS AND METHODS: Ten different metallic stents (made of steel, nitinol, tantalum, or cobalt) were implanted in plastic tubes (8 mm). The stent lumina were partially obstructed by wax (CT density, -30 H) resulting in 50-60% in-stent stenoses. The tubes were filled with diluted contrast material (25 mmol/L of gadopentetate dimeglumine or 6 mg I/mL of iodinated contrast material) and placed in a plastic container filled with oil or water, respectively. CT angiography was performed on a four-detector CT scanner (detector collimation, 4 x 1 mm; slice thickness, 1.25 mm; table feed, 4 mm per rotation). MR angiography was performed on a 1.5-T system with a three-dimensional gradient-echo sequence (TR/TE, 4.6/1.8; flip angle, 30 degrees; slice thickness, 1.88 mm). Axial and longitudinal reformations of CT and MR imaging data were evaluated regarding the in-stent attenuation and signal intensity, the visible lumen diameter inside the stent, and the delineation of the stenoses. For comparison, digital radiographic angiography was performed as the gold standard. RESULTS: The degree and character of stent-related artifacts differed in CT angiography and MR angiography. In CT angiography, only the tantalum stent caused artifacts that obscured the stenosis; in all other cases, the stenoses were visible. In MR angiography, depiction of stenoses was impaired in two steel stents but possible in the tantalum and most nitinol stents. CONCLUSION: CT angiography is suited for detection of relevant stenoses in steel, cobalt-based, and nitinol stents. MR angiography is superior only in tantalum products.     

64-slice computed tomography assessment of coronary artery stents: a phantom study

Purpose: To compare the use of a new 64-slice computed tomography (CT) scanner with 16-slice CT in the visualization of coronary artery stent lumen.

Material and Methods: Eight different coronary artery stents, each with a diameter of 3 mm, were placed in a static chest phantom. The phantom was positioned in the CT gantry at an angle of 0° and 45° towards the z-axis and examined with both a 64-slice and a 16-slice CT scanner. Effective slice thickness was 0.6 mm with 64-slice CT and 1 mm with 16-slice CT. A reconstruction increment of 0.3 mm was applied in both scanners. Image quality was assessed visually using a 5-point grading scale. Stent diameters were measured and compared using paired Wilcoxon tests.

Results: Artificial lumen reduction was significantly less with 64-slice than with 16-slice CT. Average visible stent lumen was 53.4% using 64-slice CT and 47.5% with 16-slice MSCT. Most severe artifacts were seen in stents with radiopaque markers. Using 64-slice CT, image noise increased by approximately 30% due to thinner slice thickness.

Conclusion: Improved spatial resolution of 64-slice CT resulted in superior assessment of coronary artery stent lumen compared to 16-slice CT. However, a relevant part of the stent lumen is still not assessable with multi-slice CT.     

In vitro evaluation of 56 coronary artery stents by 256-slice multi-detector coronary CT

Objective

We sought to investigate stent lumen visibility of 56 coronary stents with the newest 256-multi-slice-CT (256-MDCT) technology for different reconstruction algorithms in an in vitro model.

Background

Early identification of in-stent restenosis (ISR) is important to avoid recurrent ischemia and prevent acute myocardial infarction (AMI). Since angiography has the disadvantage of high costs and its invasiveness, MDCT could be a convenient and safe non-invasive alternative for detection of ISR.

Material and methods

Percentages of in-stent lumen diameter and in-stent signal attenuation (measured as contrast-to-noise ratio (CNR)) of 56 coronary stents (group A ≤2.5 mm; group B = 2.75–3.0 mm; group C = 3.5–4.0 mm) were evaluated in a coronary vessel in vitro phantom (iodine-filled plastic tubes) employing four different reconstruction algorithms (XCD, CC, CD, XCB) on a novel 256-MDCT (Philips-iCT, collimation = 128 mm × 0.625 mm; rotation time = 270 ms; tube current = 800 mA s with 120 kV). Analysis was conducted with the semi-automatical full-width-at-half-maximum (FWHM) method. P-values <0.05 were regarded statistically significant.

Results

In-stent lumen diameter >60% for group C stents was significantly larger and CNR was significantly lower (both p < 0.05) for sharp kernels (CD; XCD) when compared to groups A/B. The FWHM-method showed significantly smaller in-stent lumen diameter (p < 0.05) when compared to the manual method.

Conclusion

256-MDCT could potentially be employed for clinical assessment of stent patency in stents >3.0 mm when analysed with cardio-dedicated sharp kernels, although clinical studies corroborating this claim should be performed. However, stents ≤3.0 mm reconstructed by soft kernels revealed insufficient in-stent lumen visualisation and should not be used in clinical practice.Further improvements in spatial and temporal image resolution as well as reductions of radiation exposure and image noise have to be accomplished for the ambitious goal of characterising both CT coronary artery anatomy and in-stent lumen.     

冠状动脉支架内腔的可视度:64层螺旋CT与宝石CT的比较

目的评价宝石CT(HDCT)显示冠状动脉支架的能力,并与64层螺旋CT进行比较。方法 52例患者(共放置85个支架)进行64层螺旋CT和HDCT扫描,分别测量支架内CT值和相邻未放置支架节段的CT值,同时测量支架内径,并对两者进行比较。结果在HDCT上,支架内腔的显示率明显高于64层螺旋CT(分别为69.0%和42.9%,P<0.01)。支架内CT值的上升值分别为15.2 HU和54.2 HU(P<0.01)。结论由于空间分辨率的提高以及有效抑制线束硬化伪影,HDCT与64层螺旋CT比较,提高了支架内腔的可视度,能更准确测量支架的内径。     

Monoenergetic reconstructions for imaging of coronary artery stents using spectral detector CT: In-vitro experience and comparison to conventional images

Purpose

Accurate assessment of coronary stents using non-invasive CT imaging remains challenging despite new stent materials and improvements in CT technology. Virtual monoenergetic (monoE) images reconstructed from dual energy CT acquisitions potentially decrease artifacts caused by coronary stents. A novel spectral detector technology provides monoE and conventional images simultaneously for all conducted scans. The purpose of our study was to systematically investigate the influence of different monoE reconstructions on the visualization of coronary stent lumen in comparison to conventional images.

Prospective ECG-triggered axial CT at 140-kV tube voltage improves coronary in-stent restenosis visibility at a lower radiation dose compared with conventional retrospective ECG-gated helical CT

The purpose of this study was to compare coronary 64-slice CT angiography (CTA) protocols, specifically prospective electrocardiograph (ECG)-triggered and retrospective ECG-gated CT acquisition performed using a tube voltage of 140 kV and 120 kV, regarding intracoronary stent imaging. Coronary artery stents (n?=?12) with artificial in-stent restenosis (50% luminal reduction, 40 HU) on a cardiac phantom were examined by CT at heart rates of 50–75 beats per minute (bpm). The subjective visibility of in-stent restenosis was evaluated with a three-point scale (1 clearly visible, 2 visible, and 3 not visible), and artificial lumen narrowing [(inner stent diameter???measured lumen diameter)/inner stent diameter], lumen attenuation increase ratio [(in-stent attenuation???coronary lumen attenuation)/coronary lumen attenuation], and signal-to-noise ratio of in-stent lumen were determined. The effective dose was estimated. The artificial lumen narrowing (mean 43%), the increase of lumen attenuation (mean 46%), and signal-to-noise ratio (mean 7.8) were not different between CT acquisitions (p?=?0.12–0.91). However, the visibility scores of in-stent restenosis were different (p?<?0.05) between ECG-gated CTA techniques: (a) 140-kV prospective (effective dose 4.6 mSv), 1.6; (b) 120-kV prospective (3.3 mSv), 1.8; (c) 140-kV retrospective (16.4–18.8 mSv), 1.9; and (d) 120-kV retrospective (11.0–13.4 mSv), 1.9. Thus, 140-kV prospective ECG-triggered CTA improves coronary in-stent restenosis visibility at a lower radiation dose compared with retrospective ECG-gated CTA.     

Performance of dual source versus 256-slice multi-slice CT in the evaluation of 16 coronary artery stents

Introduction

Invasive coronary angiography is the reference method for identification of in-stent restenosis (ISR) bearing the disadvantages of high costs and invasiveness. New approaches like dual-source CT (DSCT) and 256-multi-slice CT (256-MSCT) may potentially be the future methods of choice to reliably exclude ISR in patients with low or intermediate risk of restenosis.We sought to compare the performance of DSCT and 256-MSCT for the in vitro assessment of stent lumen diameter and basic scan parameters in stents of various diameters and designs.

Materials and Methods

In 16 coronary artery stents we evaluated relative in-stent lumen diameter, attenuation, noise, attenuation-/signal-to-noise ratio (ANR/SNR) and radiation dose (CTDI_vol) in an acknowledged coronary vessel in vitro phantom (iodine-filled plastic tubes) with DSCT (Siemens, SOMATOM Definition, collimation = 2 × 64 × 0.6 mm, pitch = 0.26, current = 400 mAs/rot, voltage = 120 kV, tube-rotation-time = 330 ms) and 256-MSCT (Philips Brilliance, iCT, tube collimation = 2 × 128 × 0.625 mm, pitch = 0.18, current = 800 mAs_eff, voltage = 120 kV, tube-rotation-time = 270 ms). Diameter analysis was conducted with the observer-independent full-width-at-half-maximum (FWHM) technique.

Results

DSCT and 256-MSCT revealed similar stent lumen diameters (50.7 ± 7.2% vs. 50.8 ± 7.4%, p = 0.98). Attenuation (−19 ± 25 HU vs. 54 ± 29 HU), ANR (−0.9 ± 1.2 vs. 2.9 ± 1.8) and SNR (12.1 ± 2.4 vs. 17.4 ± 1.9) were better in the DSCT (all p < 0.001) at the expense of significantly higher radiation doses (CTDI_vol = 87 vs. 51 mGy, p < 0.01). Noise was comparable (21 ± 2 HU vs. 20 ± 2 HU, p = n.s.). Only stents with a diameter >3 mm allowed sufficient stent lumen assessment in both scanners and showed a relative lumen diameter of 60–66%.

Conclusions

The measured stent lumen diameter and image noise were similar in both scanners. Yet the DSCT offered a more truthful stent lumen visualization at the cost of higher radiation dose.Applying the FWHM approach only stents with a diameter >3 mm offered sufficient stent lumen assessment.