Axial Image Quality Research Articles

3D cone-beam tomosynthesis provides axial imaging of the spine with lower radiation compared to computed tomography

Three imaging techniques were compared using porcine spines. To compare image acquisition time, radiation exposure, pedicle width measurement, assessment of screw breach, and image artifact between cone-beam tomosynthesis (CBT) single mode, CBT dual mode (stereotactic CBT), and computed tomography (CT) imaging with and without spinal implants. CT is the standard for axial imaging of orthopedic procedures. CBT technology is being developed, allowing real-time intraoperative imaging and 3D surgical guidance. CBT may deliver useful axial imaging quicker with less radiation than current technologies. Six porcine spines were instrumented with bilateral pedicle screws at six levels connected with 5.5mm rods. Dosimeters were attached to four surfaces of spines. CT, CBT single and CBT dual images were acquired pre-implant and post-implant. Image acquisition and 3D reconstruction times were recorded. Pedicle widths were measured before and after instrumentation. Screw medial breaches were graded (0: no breach, 1: < 2mm, 2: 2-4mm, 3: > 4mm). Artifact and/or distortion of each image was ranked (0 = none, 1 = mild, 2 = moderate, 3 = large). Image acquisition and reconstruction times, radiation dose, pedicle width, screw breach and artifact were compared between techniques. Total image acquisition and reconstruction times of CBT was significantly less (single: 9.9 ± 0.2s, p < 0.001; dual: 60.0 ± 8.7 s, p < 0.001) than CT (250.3 ± 36.7s). CBT had significantly less radiation exposure than CT (CT: 0.7 ± 0.1rad, single: 0.03 ± 0.02rad, dual: 0.07 ± 0.03rad; p < 0.001). No difference in pedicle width change pre-implant to post-implant was found (CT: p = 0.449, single: p = 0.430, dual: p = 0.528). Pedicle width (pre-implant: p > 0.5, post-implant: p > 0.9) and pedicle width change (p > 0.4) was similar amongst all techniques. Breach assessment was not different between groups (p = 0.257). CBT images had consistently lower artifact grades than CT. Although CBT axial image quality appeared subjectively inferior to CT, it enabled consistent assessment of pedicle width and screw breach, at half time and 10× lower radiation exposure. With continued refinements, CBT technology may allow for adequate intra-operative axial imaging using low radiation exposure.

Evaluation of a respiratory navigator-gating technique in Gd-EOB-DTPA-enhanced magnetic resonance imaging for the assessment of liver tumors

ObjectivesWe investigated the clinical usefulness of respiratory navigator-gating technique for the assessment of liver tumors in Gd-EOB-DTPA-enhanced magnetic resonance (MR) imaging. MethodsEighty patients who underwent Gd-EOB-DTPA-enhanced MR imaging to evaluate known or suspected liver tumors were enrolled. Three-dimensional spoiled gradient-recalled echo images of the liver were acquired in the hepatobiliary phase by the following three methods: breath-hold imaging, navigator-gated low-resolution imaging, and navigator-gated high-resolution imaging. Navigator-gated imaging was performed during free breathing. Spatial resolution was identical between breath-hold imaging and gated low-resolution imaging. Signal intensities in the liver, muscle, and spleen were measured in 20 patients. Image quality was visually evaluated in all 80 patients. The detection rate and lesion conspicuity were assessed for 71 malignant liver lesions identified in 29 patients. ResultsThe liver-to-muscle and liver-to-spleen signal ratios were significantly lower for gated images compared to breath-hold images. Images of acceptable quality were obtained in most patients by all three methods, and the overall image quality of axial images did not differ significantly among the imaging methods, although superior reformatted coronal images were obtained by gated high-resolution imaging. The detection rates of malignant liver lesions were similar among the three imaging methods, although lesion conspicuity was significantly better for breath-hold imaging compared to gated imaging. ConclusionsNavigator-gated imaging provided image qualities and detection rates of malignant liver lesions comparable to breath-hold imaging in Gd-EOB-DTPA-enhanced MR imaging; however, no additional benefits of high-resolution imaging were proven for lesion evaluation.

CT Angiography in Potential Living Kidney Donors: 80 kVp Versus 120 kVp

The purpose of this article is to retrospectively investigate the diagnostic accuracy, image quality, and radiation dose of renal artery CT angiography (CTA), at 80 kVp compared with 120 kVp, in adult kidney donors. CTA examinations of 258 consecutive potential kidney donors were retrospectively evaluated; 189 patients were scanned using 64-MDCT scanners (higher maximal tube current), and 69 patients were scanned using 16-MDCT scanners (lower maximal tube current). On the basis of the tube potential and scanners, the study population was divided into four groups. Qualitative and quantitative analysis include vascular attenuation measurements, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR). Volume CT dose index (CTDIvol) was recorded, and size-specific dose estimate was also estimated. Using 80 kVp for the 16-MDCT scanner, there was a 64.9% reduction in size-specific dose estimate (66.1% reduction in CTDIvol), increased noise, and tube current saturation in all cases. Axial image quality was significantly lower compared with that obtained at 120 kVp (p = 0.02), but image quality and visibility of renal artery branch order were comparable. Using 80 kVp for the 64-MDCT scanner, there was a 40.5% reduction in size-specific dose estimate (43.6% reduction in CTDIvol) and increased SNR and CNR (p < 0.001). No significant differences in 3D image quality and branch order visibility were observed. Tube current saturation was reached in 31% of cases. One hundred fifty-one patients (86 imaged at 80 kVp and 65 imaged at 120 kVp) underwent donor nephrectomy; CTA diagnostic accuracy was 100%. Renal artery CTA using 80 kVp combined with limiting the tube current results in a significant reduction in radiation dose and improved SNR and CNR, without deterioration of image quality.

Image Quality of 320–Detector Row Wide-Volume Computed Tomography With Diffuse Lung Diseases

To evaluate image quality of 320-detector row wide-volume (WV) computed tomography (CT) compared to 64-detector row helical CT from axial images and coronal multiplanar reformation (MPR). Thirty-five patients with diffuse lung diseases were scanned using both 320-detector row WV CT (coneXact and volumeXact+) and 64-detector row helical protocols. Three blinded observers evaluated dislocation and heterogeneity of normal structures on 3 MPR patterns (WV scan with coneXact, WV scan with volumeXact+, and helical scan) using a 3-point scale from 1 (severe dislocation/heterogeneity) to 3 (no dislocation/heterogeneity). They also evaluated axial images of 2 scan patterns (WV with volumeXact+ and helical) using a 5-point scale from 1 (nondiagnostic) to 5 (excellent). Statistical analyses were performed with a post hoc test, Wilcoxon signed rank test, Mann-Whitney U test, or the Kendall W test. The WV scans with the coneXact algorithm had significantly lower quality scores than the WV scans with the volumeXact+ algorithm and the helical scans (P < 0.01) with MPR. Helical scans had significantly lower quality scores than the WV scans with volumeXact+ for heterogeneity on the mediastinal window setting with MPR (P < 0.01). There were no significant differences concerning total image quality of axial images between the WV scans with the volumeXact+ algorithm and the helical scans. The overall image quality of WV scans with the volumeXact+ algorithm was almost comparable to that of the helical scans on the lung window setting, but density homogeneity with helical scans was inferior to that of the WV scans with the volumeXact+ algorithm on the mediastinal window setting with MPR.

Improving the optics of KPros

Abstract Purpose The purpose of this talk is to review the clinically relevant optical and visual requirements of keratoprostheses. Methods Modelling of designs using commerical and purpose‐written optical design software as well as clinical measurments on a small number of patients implanted with the osteo‐odonto keratprosthesis optical cylinder. Results The important visual optical proerties of any keratoprosthesis optic are the foveal image quality (affecting the acuity) and the field of view. These fundamentally depend upon the pupil size, and the length and diameter of the optic since the power must be fixed to give emmetropia. The axial image quality can be changed by controlling the spherical aberration. However, maximising axial (foveal) image quality causes poorer off‐axis image quality potentially reducing the effective visual field. Shorter, larger diameter optics help improve the visual field as well as consideration of off‐axis image quality. Conclusion The theoretical properties of keratoprosthesis optics must be balanced against other clinical requirements such as longterm complications and stability. The clinical requirements provide the constraints on the optical design. However, discussions can usefully happen between surgeons and those involved in the design of the optics to give the best possible visual rehabilitation.

Using the K-edge to Improve Contrast Conspicuity and to Lower Radiation Dose With a 16-MDCT

To study the effect of tube current (milliamperes; mA) and potential (peak kilovolt; kVp) on the attenuation values and contrast-to-noise ratios of iodine- and gadolinium-based contrast media (CM) for computed tomography angiography (CTA). Phantom Study: A water-filled phantom with five 20-mL syringes filled with 1:20 dilution of 282, 300, 370, and 400 milligrams of iodine per milliliter concentration CM and gadopentetate dimeglumine (Magnevist, Berlex Laboratories, Wayne, NJ, 0.5 mol/L) was scanned with a 16-multidetector CT using 80, 100, 120, and 140 kVp and 500-millisecond gantry rotation time. The milliampere was either fixed at 100, 200, 300, and 380 or automatically adjusted with noise indices of 15, 20, and 25 or manually adjusted to maintain a constant image noise. The attenuation value (Hounsfield unit; HU) and its standard deviation of CM in each syringe and of the water phantom were obtained. Statistical analysis was performed to determine difference between attenuation values and contrast medium-to-water contrast-to-noise ratios at various kVp and mA selection. Human Study: Three groups of patients had CTA for abdominal aortic aneurysm with similar computed tomography parameters, varying only in kVp selection of either 100 (group A), 120 (group B), or 140 (group C). Another group (group D) had CTA at 100 kVp but with the CM volume reduced to 60%. The CTA studies were evaluated for the quality of axial and 3D images; attenuation values in the aorta, superior mesenteric artery, and iliac arteries; image noise; and scanner-estimated radiation dose. Statistical analysis was performed to determine the difference in image quality and radiation dose among the groups. Phantom Study: In comparison with 140 kVp, regardless of selected milliampere or noise indices, images acquired at 80, 100, and 120 kVp showed 90.8% to 94.2%, 47% to 49.7%, and 18.9% to 20.7% (P<0.0001) increase in HU of iodine-based CM, respectively, and 62.9%, 39.6%, and 16.8% (P<0.0001) increase in HU of gadolinium-based CM, respectively. Human Study: The axial images in all the groups were diagnostically acceptable. There was significantly inferior quality of axial images associated with greater image noise in group A and group D (P<0.01) in comparison with group C, but there was no difference in the quality of 3D images among the 4 groups. In comparison with group C, there was significantly higher attenuation of the aorta, superior mesenteric artery, and iliac arteries in group A (P<0.01), group B (P<0.05), and group D (P<0.01). The radiation dose (CT dose index volume) decreased to 12+/-4 in groups A and D compared with 17+/-4 in group B and 24+/-5 in group C. Lower kVp increases the attenuation of iodinated and gadolinium CM. CTA of the abdominal aorta performed at low kVp results in higher attenuation in aorta with reduced radiation dose and without degrading the diagnostic image quality. The iodinated CM volume can be reduced by reducing kVp during CTA.

Comparison of retrospectively ECG-gated and nongated MDCT of the chest in an emergency setting regarding workflow, image quality, and diagnostic certainty

This study aims to assess the influence of ECG-gated acquisition on workflow and to compare image quality and diagnostic certainty for retrospectively ECG-gated and nongated multidetector computed tomography of the chest in the emergency suite. Thirty-two consecutive patients were referred for both an ECG-gated and a nongated CT to rule out traumatic thoracic injury (n=15) or acute aortic dissection (n=17). The time from the start of the transportation from the emergency suite to the CT room until the start of the CT scan was recorded. Using a scoring system, the image quality of axial images and multiplanar reformats, the presence of disease, and the subjective diagnostic certainty were assessed with regard to the vascular structures, the bone structures, and the lung parenchyma. The time needed for transportation and patient preparation was 12.1+/-1.7 min (8.1-14.5 min). The motion artifacts of the thoracic aorta and the supra-aortic vessels were significantly reduced in the ECG-gated data acquisition compared with the nongated technique (P<0.001). Subjective diagnostic certainty for assessment of the aorta was significantly better using ECG gating. The image quality of the lung parenchyma (P<0.005), the spine (P<0.005), and the ribs (P<0.002) was inferior in the ECG-gated data sets but did not compromise the detection rate of traumatic lesions and fractures. Performing ECG gating in the emergency room did not slow down the diagnostic workup. ECG-gated acquisition performed better in the assessment of the aorta, but image quality for lung and bone structures was slightly reduced. Further studies are required to assess the influence of the imaging technique on the diagnostic outcome.

Theoretical analysis of peripheral imaging after excimer laser corneal refractive surgery for myopia

Purpose To explore theoretically the retinal point images in the peripheral fields of eyes that have had excimer laser refractive surgery. Setting University research laboratory. Methods Model eyes were based on Navarro’s finite schematic eye, the eyes being made myopic by an increase in axial length. To simulate photorefractive keratectomy (PRK), the anterior shape and thickness of the cornea were modified. Variables included pupil size, ablation zone size, preexisting refractive error, and the addition of a blending zone. Image-quality criteria for each retinal point image were its size and the angular separation of the centroids of those parts of the image produced by rays passing through ablated and unablated corneal zones. Results In the peripheral visual field, the boundary between the ablated and unablated cornea caused a separation of the retinal image of a single point into 2 parts. The separation increased with the preexisting refractive error. Image quality was correspondingly reduced by ablation. As pupil size increased, the field angle at which the retinal image doubling first occurred decreased. Increasing the diameter of the ablation zone or using a blending zone increased the angle at which the doubling first occurred, and the blending zone improved image quality considerably. Chromatic effects appeared to be relatively unimportant. Conclusions This analysis provides further evidence of the disadvantages of small central ablation zones in excimer laser refractive surgery and of the advantages of well-designed blending zones in improving postsurgical peripheral image quality. Image quality in the peripheral field of the pseudoemmetropic post-PRK eye is generally worse than in a naturally emmetropic eye, even though the axial image quality may be similar.

Multiple point sources collimated with transaxial septa for high energy gamma camera transmission scanning

The axial imaging properties of the transmission scanning system for attenuation correction of gamma camera coincidence imaging are evaluated. This system uses multiple Cs-137 point sources collimated by transaxial septa already in place for emission imaging. Both the spacing and the degree of collimation of the sources can be controlled. Axial image quality was evaluated with three levels of collimation (4 mm, 7 mm, and 10 mm slot, all 6 cm deep) and two source spacing distances (1.3 cm and 2.6 cm). A multiple disk phantom with adjustable spacing was imaged, as well as a thorax phantom. The 4 mm collimation and 1.3 cm source spacing provided well-resolved images of the disks with 2.9 cm center-to-center spacing. The best collimation/spacing combinations come with substantial overlap of flux from difference sources in the detector. Axial resolution can be adequately controlled with extra lead collimation of the sources.

Image quality in polypseudophakia for extremely short eyes

AIMTo evaluate the image quality produced by polypseudophakia used for strongly hypermetropic and nanophthalmic eyes.METHODSPrimary aberration theory and ray tracing analysis were used to calculate the optimum lens shapes and...

Cervical spine MR imaging: generating high-signal CSF in sagittal and axial images.

Three magnetic resonance (MR) imaging techniques were compared as to their ability to generate images with high-signal cerebrospinal fluid (CSF) to provide a high-contrast CSF-dura interface. The three techniques were CSF gating to the peripheral pulse, selective saturation recovery with gradient refocusing (SSRGR), and gradient recalled acquisition in the steady state (GRASS). In sagittal views of the cervical spine, CSF gating proved to be a reliable technique for producing images with uniform high-signal CSF in a single-section or multi-section mode. In axial views, SSRGR and GRASS techniques were more consistent than CSF gating in producing high-signal CSF images, especially in a multisection mode. Although axial image quality was nearly equivalent for SSRGR and GRASS techniques, the latter was clinically more efficient because of shorter imaging times. These methods of imaging in the cervical spine yield sufficient CSF-dura contrast and spatial resolution to be of use in the diagnosis of cervical disk disease.