Total digital radiology department: spatial resolution requirements

The minimum spatial resolution required for a total digital radiology department has yet to be defined. A pilot study designed to provide this information was performed. Abnormal and normal radiographic images of children were digitized and redisplayed on film at spatial resolutions of 5.0, 2.5, 1.25, and 0.625 lp/mm. These resolutions are comparable to a digital display of a 14 X 14 in. chest image having pixel elements of 4096 X 4096, 2048 X 2048, 1024 X 1024, and 512 X 512, respectively. Contrast resolution was maintained at 12 bits or 4096 gray levels. The three phases of data acquisition were (1) the standard analysis of receiver operating characteristics, (2) a checklist evaluation of the "seeability" of important structures, and (3) a comparison of all resolutions and a discernment of usability. Fifteen radiologists participated in the study. On the basis of the pediatric cases used, the results showed that the needed spatial resolution for a total digital radiology department may be around 2.5 lp/mm (2048 X 2048). Checklist data on seeability of structures and comparisons of all resolutions give information on specific changes that are occurring as the resolution is decreased, and, when included with the receiver-operating-characteristic data, they become a major component in developing a resolution standard. The finding that 2.5 lp/mm is the required spatial resolution makes construction of a total digital radiology department possible with present state-of-the-art technology.     

Blood flow determinations utilizing digital densitometry

A method of obtaining relative and absolute blood flow measurements from digital densitometry was evaluated with a simulated vessel phantom and a hydrodynamic model. A digital vascular imaging system capable of acquisition in 512(2) and 1024(2) mode was used. Relative and absolute blood flow were measured using parameters derived from the densitometric curve. Since application of densitometric data to absolute flow measurements requires the vessel diameter, an algorithm for vessel size determination was created. Gray scale changes were demonstrated to be linearly related to contrast concentration. The variance of vessel size determination was significantly different in all combinations of 1024(2) and 512(2) imaging with 15 cm or 35 cm field size. The error in vessel size determination was significantly less using the larger 1024(2) matrix and the smaller 15 cm image intensifier field size, as shown by the smaller variance. In relative flow determinations, there was good correlation between the flow and four parameters of the densitometric curve with no significant differences between 512(2) and 1024(2) imaging. Absolute flow determinations had slightly lower correlation to actual flow but were not significantly different from relative flow determinations. Relative and absolute blood flow determinations can be performed adequately with either 512(2) or 1024(2) imaging. The increased accuracy in vessel size determination with 1024(2) imaging makes this high resolution system potentially preferable to determine absolute blood flow.     

Quantitative coronary arteriography: design and validation

The authors assessed the performance of an automatic and rapid coronary quantification method by evaluating its accuracy in a stenosis phantom. Measurements were obtained with a lucite phantom with 2-, 3-, and 4-mm vessel diameters and concentric stenoses of 33%, 50%, 67%, and 75%. Direct digital angiographic images as well as 10 X 10 spot films and 35-mm cine angiography films were acquired with and without structural noise and mask subtraction. The films were digitized with magnification factors of one and two. An interactive analysis program was used to automatically determine the vessel edges with a Gaussian fit to the cross-sectional density profiles perpendicular to the center line of the vessel. Relative changes of the densitometric cross-sectional area along the vessel were used to assess the percentage of stenosis. Densitometric measurements were comparable in both digital and cine angiograms (r = .99 and r = .98, respectively); however, diameter measurements showed a higher variability and were dependent on the amount of magnification applied to the images.     

Quantitative evaluation of measurement accuracy for three-dimensional angiography system using various phantoms

PURPOSE: The purpose of this study was to evaluate the spatial resolution and accuracy of three-dimensional (3D) distance measurements performed with 3D angiography using various phantoms. MATERIALS AND METHODS: With a 3D angiography system, digital images with a 512 x 512 matrix were obtained with the C-arm sweep, which rotates at a speed of 30 degrees/second. A 3D comb phantom was designed to assess spatial resolution and artifacts at 3D angiography and consisted of six combs with different pitches: 0.5 mm, 0.6 mm, 0.7 mm, 0.8 mm, 0.9 mm, and 1.0 mm. Frame rate, field of view (FOV) size, reconstruction matrix, and direction of the phantom were changed. In order to investigate the accuracy of 3D distance measurements, aneurysm phantoms and stenosis phantoms were used. Aneurysm phantoms simulated intracranial saccular aneurysms and parent arteries; 2-mm- or 4-mm-inner-diameter cylinder and five different spheres (diameter: 10, 7, 5, 3, 2 mm) were used. Stenosis phantoms were designed to simulate intracranial steno-occlusive diseases; the nonpulsatile phantoms were made of four cylinders (diameter: 3.0, 3.6, 4.0, 5.0 mm) that had areas of 50% and 75% stenosis. The dimensions of the spheres and cylinders were measured on magnified multiplanar reconstruction (MPR) images. RESULTS: The pitch of the 0.5 mm comb phantom was identified clearly on 3D images reconstructed with a frame rate of 30 frame/sec and 512(3) reconstruction mode. In any reconstruction matrixes and any angles of the phantom, the resolution and artifacts worsened when frame rates were decreased. With regard to the angle of the phantom to the axis of rotational angiography, spatial resolution and artifacts worsened with increase in angle. Spatial resolution and artifacts were better with a FOV of 7 x 7 inch than with one of 9 x 9 inch. All spheres on the aneurysm phantom were clearly demonstrated at any angle; measurement error of sphere size was 0.3 mm or less for 512(3) reconstruction. In 512(3) reconstruction, the error of percent stenosis was 3% or less except for a cylinder diameter of 3.0 mm and 5% for a cylinder diameter of 3.0 mm. CONCLUSION: Spatial resolution of the reconstructed 3D images in this system was 0.5 mm or less. Measurement error of sphere size was 0.3 mm or less when 512(3) reconstruction was used. When using proper imaging parameters and postprocessing methods, measurements of aneurysm size and percent stenosis on the reconstructed 3D angiograms were substantially reliable.     

Comparison of observer and videodensitometric measurements of simulated coronary artery stenoses

To test the applicability of an automated vessel measurement technique to coronary arteriography, a videodensitometric method with caliper measurements on digital subtraction images of a moving coronary artery phantom was compared. Percent diameter stenosis was determined by both methods, revealing a twofold improvement in reproducibility with the videodensitometric method, with percent stenosis being determined within +/- 10% for two different iodine concentrations injected during continuous flow into the simulated coronary arteries. Absolute diameters were also measured by the videodensitometric method, showing a high degree of correlation between measured and true diameter for vessels between 0.5-3.0 mm.     

Comparison of 2048-line digital display formats and conventional radiographs: an ROC study

Observer performance tests were conducted to compare the effects on diagnostic accuracy of digital hard copy and video display formats versus conventional radiographic film. Digital images were obtained by digitizing conventional chest radiographs to a 2048 x 2048 matrix with a laser film scanner. Three digital display formats were used: laser-printed digital film, a 2048-line video monitor without user interaction, and a 2048-line video monitor with user interaction. Thirty-one posteroanterior chest radiographs, determined by consensus of four thoracic radiologists to contain septal lines (n = 11), parenchymal nodules (n = 7), nodules and septal lines (n = 7), or neither abnormality (n = 6), were used for the study. Images were interpreted by four radiologists in four separate viewing sessions. Diagnostic accuracy was determined by receiver-operating characteristic analysis for each observer with each viewing technique. No statistical differences in diagnostic accuracy, determined by the area under the receiver-operating-characteristic curve, were found between the analog film, the digital film, and the two video digital display formats. This preliminary study suggests that 2048-line digital displays may be an acceptable alternative to the traditional lightbox viewing method for the perception of these two abnormalities commonly seen on chest radiographs.     

Effect of sharpening filter on vessel diameter measured by quantitative coronary arteriography

At cardiac catheterization, analog images obtained using cinefilm are translated into digital images, and images appearing on the CRT are filtered by a sharpen filter. We investigated the effect of the sharpening filter on vessel diameter as measured by quantitative coronary arteriography. We acquired images of a vessel phantom filled with contrast material using an X-ray image intensifier. Vessel diameters measured by quantitative coronary arteriography were 1 mm, 1.5 mm, 2 mm, 3 mm, 4 mm, and 6 mm. Results showed that vessel diameters were decreased when the sharpening filter was used and that more intense filtering decreased the measured diameter further. When the diameter of the vessel phantom was less than 4 mm, the diameter was smaller and the ratio of decrease was larger. Vessel diameters of 2 mm, 3 mm, 4 mm, and 6 mm measured a maximum of 2.9 smaller, while those of 1 mm and 1.5 mm measured a maximum of 9.2 smaller.     

Videodensitometric analysis of coronary stenoses. In vivo geometric and physiologic validation in humans

Assessment of the severity of coronary stenoses on arteriograms conventionally is based on subjective estimates of percent luminal diameter narrowing. However, in studies in patients with multivessel coronary artery disease, we have found a poor correlation between percent stenosis and the physiologic significance of an individual coronary obstruction. The purpose of this study was to determine whether computerized videodensitometry would allow estimation of coronary luminal area and therefore prediction of the physiologic significance of individual coronary stenoses in humans. Videodensitometry was used to define the minimal luminal area of 15 left anterior descending, 15 circumflex, and 15 right coronary artery segments in 43 patients. Computer-assisted quantitative coronary arteriography (method of Brown et al) was used to determine the minimal luminal cross-sectional area of these same segments. In each arterial segment, coronary vasodilator reserve was assessed using intraoperative (n = 18 segments) or intracoronary (n = 27 segments) Doppler measurements of coronary vasodilator reserve. Videodensitometric estimates of coronary luminal area correlated well with minimal luminal area defined using the independent geometric technique of quantitative coronary arteriography (r = 0.82, y = 0.97 X + 0.71, SEE = 1.83 mm2, n = 45) and with lesion physiologic significance as defined by studies of the peak-to-resting velocity ratio (r = 0.71, 0.92, and 0.74 for the left anterior descending, circumflex, and right coronary arteries, respectively). Thus, videodensitometry is a promising method that may supplement geometric approaches to quantitative analysis of coronary arteriograms in humans.     

Accuracy of quantitative coronary angiography for different directions of the coronary artery

We evaluated the effect of changes in the direction of the coronary artery in terms of the accuracy and precision of vessel diameter measurement in a quantitative coronary angiography system (QCA system). Vessel phantoms sized 0.3, 0.5, 1.0, 1.5, 2.0, and 2.5 mm in diameter were evaluated. The phantoms were aligned on an acrylic plate, and the angle to the television (TV) camera was altered. The deployed angles were 0 (perpendicular), 45, 90, and 135 degrees in clockwise order. The phantoms were imaged with matrices of 1024 x 1024 (1024(2)), 512 x 512 (512(2)), and 512 x 1024. Image size was 7 inches, and the frame rate was 15 frames per second. Minimal lumen diameters were measured on the ACA system. The results revealed that, in the 1024(2) matrix, overall accuracy for the 90-degree angle was significantly underestimated compared with the 0-degree angle (-0.14 vs. -0.014 mm; p=0.007). Accuracy for the 90-degree angle was better than that for the 0-degree angle when the vessel diameter was 1 mm or smaller (-0.02+/-0.16 vs. 0.10+/-0.22 mm). In addition, precision was better at the 90-degree angle than with the other angles in the 1024(2) matrix (overall precision=0.002 mm). In the 512(2) matrix, overall accuracy for the 90-degree angle was significantly underestimated compared with the 45-degree angle (-0.077 vs. 0.096 mm; p=0.02). In addition, accuracy for the 90-degree angle was better than that for the 45-degree angle below 1 mm (0.05+/-0.24 mm vs. 0.26+/-0.47 mm). In terms of overall accuracy, the 45-degree angle in the 512(2) matrix showed significant overestimation compared with that in the 1024(2) matrix (0.096 vs. -0.069 mm; p=0.015). There was no difference in accuracy in the 512 x 1024 matrix. Our results suggest that the direction of the vessel against the TV image affects accuracy of measurement in the QCA system.     

Digital subtraction angiography: a comparison of 512(2) and 1024(2) imaging

A commercial DSA unit was modified by the manufacturer to permit 1024 X 1024 8-bit imaging. System upgrade includes a high-resolution 1049-line TV camera that operates with variable aperture to minimize x-ray exposure during 1024(2) imaging. To compare the change in resolution and radiation exposure between 512(2) and 1024(2) imaging with this system, a two-phase phantom study was performed using a high-contrast converging lead line phantom and a specially designed high-resolution low-contrast Lucite phantom. The two-phase phantom study tested general system resolution performance and resolution under simulated and actual clinical conditions for each field size (15, 25, and 36 cm). The 512(2) imaging was performed with the aperture reduced to the 512 setting; 1024(2) imaging was performed with the aperture at the 512 and 1024 values. The 1024(2) imaging resulted in only modest improvement in resolution compared to 512(2). While Nyquist limits were approached with 512(2) imaging, this was not the case with 1024(2) imaging. This suggests other factors such as system noise are playing a significant role in 1024(2) image degradation.     

Absolute cross-sectional area measurements in quantitative coronary arteriography by dual-energy DSA

Recent studies have emphasized the limitations of conventional coronary angiography. These limitations include the lack of correlation between the severity of coronary stenosis as estimated from coronary angiograms and the actual severity of stenotic lesions measured in postmortem hearts. As a result, attempts have been made to quantitate luminal dimension more precisely. The application of quantitative digital subtraction angiography (DSA) in the assessment of coronary artery lesion dimension has been limited by cardiac and respiratory motion artifacts. We have reported previously on a motion-immune dual-energy (DE) cardiac mode in which kVp and filtration are switched at 30 Hz. To assess the potential advantages of a videodensitometric technique for quantification of absolute vessel cross-sectional area (CSA), three different quantitative coronary arteriography (QCA) algorithms were compared. The three algorithms under comparison were a videodensitometric (V) algorithm, which does not require any geometric assumption for absolute vessel CSA measurement, and videodensitometric (VC) and edge detection (ED) algorithms, which do require the assumption of circular cross-section for CSA measurements. A cylindrical vessel phantom (0.5-4.75 mm in diameter) and a crescentic vessel phantom, producing 25% to 90% area stenosis, were imaged over the chest of a humanoid phantom. The low- and high-energy images were corrected for scatter and veiling glare before energy subtraction. For CSA measurements in crescentic vessel phantoms, the V algorithm produced significantly improved results (slope = 0.87, intercept = 0.51 mm2, r = .95) when compared to the VC (slope = 1.05, intercept = 4.19 mm2, r = .75) and the ED (slope = 1.57, intercept = 5.21 mm2, r = .60) algorithms.     

The effect of scatter on densitometric stenosis measurements using an experimental digital subtraction angiography system

Purpose: To examine the effect of changing the experimental scatter condition on densitometric stenosis measurements.Methods: Experiments were conducted using an experimental digital fluoroscopic system. A test phantom with a known degree of stenosis was imaged before and after the introduction of an iodine-based contrast medium to simulate the digital subtraction angiography (DSA) process. The scatter to primary ratio has been reported to be affected by area irradiated, thickness of the part, kVp, and use of scatter reduction methods such as grids and airgaps. Each of these conditions were manipulated experimentally with densitometric stenosis measurements determined for each scatter condition.Results: Theoretically and experimentally, it is shown that the densitometric analysis produces a grossly inaccurate measurement of the degree of vessel stenosis when performed on images acquired under scatter conditions.     

High-spatial-resolution MR angiography of renal arteries with integrated parallel acquisitions: comparison with digital subtraction angiography and US

PURPOSE: To retrospectively compare three-dimensional gadolinium-enhanced magnetic resonance (MR) angiography, performed with an integrated parallel acquisition technique for high isotropic spatial resolution, with selective digital subtraction angiography (DSA) and intravascular ultrasonography (US) for accuracy of diameter and area measurements in renal artery stenosis. MATERIALS AND METHODS: The study was approved by the institutional review board, and consent was obtained from all patients. Forty-five patients (17 women, 28 men; mean age, 62.2 years) were evaluated for suspected renal artery stenosis. Three-dimensional gadolinium-enhanced MR angiograms were acquired with isotropic spatial resolution of 0.8 x 0.8 x 0.9 mm in 23-second breath-hold with an integrated parallel acquisition technique. In-plane diameter of stenosis was measured along vessel axis, and perpendicular diameter and area of stenosis were assessed in cross sections orthogonal to vessel axis, on multiplanar reformations. Interobserver agreement between two radiologists in measurements of in-plane and perpendicular diameters of stenosis and perpendicular area of stenosis was assessed with mean percentage of difference. In a subset of patients, degree of stenosis at MR angiography was compared with that at DSA (n = 20) and intravascular US (n = 11) by using Bland-Altman plots and correlation analyses. RESULTS: Mean percentage of difference in stenosis measurement was reduced from 39.3% +/- 78.4 (standard deviation) with use of in-plane views to 12.6% +/- 9.5 with use of cross-sectional views (P < .05). Interobserver agreement for stenosis grading based on perpendicular area of stenosis was significantly better than that for stenosis grading based on in-plane diameter of stenosis (mean percentage of difference, 15.2% +/- 24.2 vs 54.9% +/- 186.9; P < .001). Measurements of perpendicular area of stenosis on MR angiograms correlated well with those on intravascular US images (r(2) = 0.90). CONCLUSION: Evaluation of cross-sectional images reconstructed from high-spatial-resolution three-dimensional gadolinium-enhanced MR renal angiographic data increases the accuracy of the technique and decreases interobserver variability.     

Vessel diameter measurement using digital subtraction radiography

A method for obtaining absolute diameter and cross-sectional area measurements on subtraction digital images is described and tested in phantom vessels from 1.5 to 5.5 mm in diameter filled with iodine contrast at concentrations from 23 to 185 mg I/ml. A highly linear correlation of true vs. calculated diameter is demonstrated, with accuracy and reproducibility of the method varying from +/- 1% to 2% at the highest iodine concentration to +/- 30% in the smallest tube at the lowest concentration. A method is described for correction of the observed video density values to allow for nonlinearity of response of the imaging system to iodine density, and its effect on the measured diameters is demonstrated.     

Arterial digital subtraction angiography of the lower extremities using a high-resolution 1024 image matrix

The quality of radiographs representing corkscrew vessels and collateral vessels that developed after occlusion of the superior femoral artery, using digital angiography technique with a 1024 matrix, is compared with that of a 512 matrix. The images obtained with the 1024 matrix were superior to those obtained with a 512 matrix. Using a small cannula for the application of the contrast agent, the quality of the images representing the collateral vessels was very good in 83% of the images produced with the 1024 matrix and in 70% of the 512 matrix images.     

Digital angiography: current status

Digital angiography depends upon computer recording and manipulation of the radiographic image rather than the use of film. We use a small focal spot x-ray tube, a cesium iodide image intensifier and the image is recorded on a 512 X 512 matrix at 1.2 frames per second. Subtraction angiograms have been obtained in 250 patients and 86% of the results were graded as either good or excellent in head and neck examinations. There is however a small proportion of patients in whom results are unsatisfactory and in whom this fact can be attributed most often to cardiac failure. The advantages of digital intravenous and intra-arterial subtraction angiography have been evaluated and amount to improved patient safety, high system contrast and favourable cost effectiveness; these outweigh the limitations of the restricted field of view and the relatively low resolution of the image in comparison with conventional arteriography.     

Intravenous digital subtraction angiography of arteriosclerotic vertebrobasilar disease

Intravenous digital subtraction angiography (DSA) was performed in 111 patients with vertebrobasilar ischemia. Ninety percent of the vertebral images were of diagnostic quality; 23% of the basilar images were good quality and 53% fair quality; and 58% of the posterior cerebral images were poor. Compared with selective film arteriography in 23 patients, DSA tended to underestimate the degree of atheromatous disease. Segments of the basilar artery were often poorly seen, which could result in false-negative errors. DSA can provide a general assessment of atheromatous disease of the brachiocephalic vessels, including the vertebral and carotid arteries, and in many cases can exclude occlusion or critical stenosis of the vertebrobasilar system. However, it does not adequately image the posterior cerebral or cerebellar artery.     

Intraarterial digital subtraction arteriographic evaluation of extremity tumors: comparison with conventional arteriography

Conventional arteriography and intraarterial digital subtraction arteriography (IADSA) were compared in 36 patients with primary bone or soft-tissue tumors of the extremities. The sensitivity of IADSA was at least equal to conventional arteriography for demonstrating normal or abnormal major arteries and feeding arteries, equal to or superior for depicting tumor stains or draining veins, but slightly inferior for revealing minute tumor vessels. An increase of the matrix size from 256 X 256 to 512 X 512 improved these sensitivities. IADSA with 15% diatrizoate contrast material eliminated the contrast material-induced pain in all patients. With a computer-controlled iris setting, an average of 5 minutes of procedure time and 1.7 R of radiation (0.44 mC kg) per examination could be saved. IADSA reduced the cost of an examination by an average of $67. The results indicate that IADSA was diagnostic in all instances and can replace conventional arteriography for the evaluation of extremity tumors.     

Digital subtraction angiography with an Isocon camera system: clinical applications

A new imaging system for digital subtraction angiography (DSA) was evaluated in 30 clinical studies. The image receptor is a 25 X 25 cm, 12 par gadolinium oxysulfate rare-earth screen whose light output is focused to a low-light-level Isocon camera. The video signal is digitized and processed by an image-array processor containing 31 512 X 512 memories 8 bits deep. In most patients, intraarterial DSA studies were done in conjunction with conventional arteriography. In these arterial studies, images adequate to make a specific diagnosis were obtained using half the radiation dose and half the amount of contrast material needed for conventional angiography. In eight intravenous studies performed either to identify renal artery stenosis or for evaluation of congenital heart anomalies, the images were diagnostic but objectionably noisy.     

Pseudostenosis in vessels adjacent to intracranial aneurysms on volume-rendered 3D angiograms: a phantom study

RATIONALE AND OBJECTIVE: Among artifacts on three-dimensional (3D) angiograms, pseudostenosis in vessels adjacent to intracranial aneurysms has not been described. By using a phantom, artifacts seen in vessels adjacent to intracranial aneurysms on volume-rendered 3D angiograms were assessed. MATERIALS AND METHODS: Using a 3D angiography system and a C-arm sweep, digital images were obtained with a 512 x 512 matrix. Rotation was 30 degrees /second, and frame rate was 30 frames/second. Phantom aneurysms were designed to simulate intracranial saccular aneurysms and their parent arteries. Phantoms, consisting of a cylinder (inner diameter, 2 or 4 mm) and spheres, 10, 7, 5, 3, or 2 mm in diameter, were placed at 0 degrees and 45 degrees to the axis of rotation. Two radiologists consensually recorded their findings regarding the presence and location of stenosis and its relationship to the angle of rotation. The maximum percentage of stenosis in the pseudostenosis area was measured on multiplanar reconstruction images by using a workstation computer. RESULTS: Pseudostenosis was observed in the cylinder adjacent to the sphere at both 0 degrees and 45 degrees angles; it was on a plane perpendicular to the axis of rotation. Pseudostenosis was most obvious with 10-mm spheres; it was not seen when spheres were 3 mm or less in diameter. The maximum percentage of stenosis of the pseudostenosis increased with sphere size. CONCLUSION: On volume-rendered 3D angiograms, pseudostenosis was seen in the cylinder adjacent to the sphere. The artifact lay on a plane perpendicular to the axis of rotation, and sphere size affected the artifact.