Accurate analysis of blood vessel sizes and stenotic lesions using stereoscopic DSA system

We have developed a technique to determine accurately the magnification factor and three-dimensional orientation of a vessel segment from a stereoscopic pair of digital subtraction angiograms (DSA). Our DSA system includes a stereoscopic x-ray tube with a 25-mm focal spot shift. The magnification and orientation of a selected vessel segment are determined from the distance and direction of the focal spot shift and the stereoscopic discrepancy in image positions for that segment. Our results indicate that the accuracies of determining the magnification and orientation are less than 1% and approximately 5 degrees, respectively. After the magnification and orientation are determined accurately, an iterative deconvolution technique for the measurement of vessel image size is applied to the selected vessel segment. This iterative deconvolution technique provides the best estimate of vessel image size by taking into account the unsharpness of the digital system. With this technique, the vessel image size can be determined to an accuracy of approximately 1.0 mm, which corresponds to one third the pixel size of our DSA system. Information derived from stereoscopic analysis and iterative deconvolution thus allows accurate calculation of actual vascular dimensions from DSA images.     

Automated tracking and computer reproduction of vessels in DSA images

We are developing an automated vessel-tracking method based on the double-square-box region-of-search technique, for efficient tracking of the connected vascular tree in a digital subtraction angiography (DSA) image. Tracking points and branch vessels are located by searching of the perimeter of boxes, which are centered on previously determined tracking points. The most accurate results (90% true-positive rate with six false-positives per image) are obtained by tracking using the double-square-box method. In relatively straight regions of vessels, a large box is employed for efficient tracking; in curved regions of vessels, a small box is employed to ensure accurate tracking. When tracking is completed, accurate vessel information, ie, the vessel position, size, and contrast determined at each tracking point, is available for further quantitative analysis. Computer reproductions of tracked vessel trees appear to correspond well to those in DSA images.     

Measurement accuracy of sonographic sector scanners

Discrepancies identified in anatomic size measurements in clinical obstetric sonographic studies prompted a study of the accuracy of measurements obtained from sonographic mechanical sector scanners. There were large variations in equipment calibration of the size of the sector scan angle among six clinical sonographic units from four manufacturers, causing transverse measurements to vary from 47% (high) to 9% (low), compared with the true size of a cylindrical plastic phantom. Depth measurements, parallel to the direction of sound penetration, are only velocity dependent and were accurate to within 6%. Transverse size measurements made using mechanical sector scanners should be suspect unless the calibration has been checked using a phantom test object.     

Automated calibration in vascular X-ray images using the accurate localization of catheter marker bands

RATIONALE AND OBJECTIVES: To develop a new automated calibration method for vessel measurements in vascular x-ray images. METHODS: Radiopaque marker bands mounted equidistantly on a small catheter were acquired in vitro at five image intensifier (II) sizes in x-ray projection images. The positions of the marker centers were detected by using a Hough transform and were computed at subpixel precision by using either a novel, iterative center-of-gravity approach (CGA) or a symmetry filter. Curve-fitting procedures were used to reject false-positive marker detections and to calculate intermarker distances. The calibration factor was calculated from the true marker distance and the average of the measured distances in pixels. Results were compared statistically with a grid calibration method, which was taken as the gold standard. A simulation study was performed to assess the influence of image noise on the CGA method. RESULTS: The iterative CGA method was convergent and faster than the symmetry-based technique. For four II sizes (17, 20, 25, and 31 cm), the results from the CGA method were not significantly different from the results obtained with grid calibration. For the II size of 38 cm, a significant difference (0.3% of the grid calibration factor) was found; however, this was caused by the quantification error in the image data and was not clinically relevant. In general, the performance of the CGA method improved with increasing signal-to-noise ratio. CONCLUSIONS: A practical new calibration method for small catheter sizes was developed and validated for quantitative vascular arteriography.     

MR angiography of the pedal arteries with gadobenate dimeglumine, a contrast agent with increased relaxivity, and comparison with selective intraarterial DSA

PURPOSE: To compare gadobenate dimeglumine (Gd-BOPTA)-enhanced MR angiography (i.e., contrast-enhanced MRA [CE-MRA]) of the pedal vasculature with selective digital subtraction angiography (DSA) in patients with peripheral arterial occlusive disease (PAOD). MATERIALS AND METHODS: A total of 22 patients with PAOD were prospectively examined at 1.5T. For contrast enhancement, 0.1 mmol/kg body weight of Gd-BOPTA were applied. MRA consisted of dynamic imaging with acquisition of six consecutive data sets. Acquisition time for each data set was 24 seconds, voxel size was 1.0 x 1.0 x 1.3 mm(3). A total of 20 out of 22 patient underwent selective DSA, two patients fine-needle DSA. DSA and MRA were performed within seven days. Image analysis was independently done by two observers with assessment of overall image quality, motion artifacts, detection of patent vessel segments of the distal calf and pedal vessels, and the number of patent metatarsal arteries. After four weeks, a consensus reading of DSA images was done. A second consensus reading of CE-MRA was performed after a further six weeks. RESULTS: Consensus readings of MRA and DSA revealed higher image quality and fewer motion artifacts for MRA (P = 0.021 and P = 0.008, respectively, sign test); interobserver agreement was good (kappa = 0.78) for image quality, and moderate (kappa = 0.46) for motion artifacts. There were no differences between CE-MRA and DSA in detecting patent vessel segments with a high degree of agreement (kappa = 0.89), and interobserver agreement for MRA was substantial (kappa = 0.89). Significantly more vessels were assessed as partially occluded on DSA than on CE-MRA (P = 0.004). There was a good agreement between DSA and CE-MRA for assessment of relevant vessel stenosis (kappa = 0.61); interobserver agreement for MRA was good (kappa = 0.65). CE-MRA detected significantly more patent metatarsal arteries than did DSA (P < 0.001). CONCLUSION: Gd-BOPTA-enhanced MRA is comparable to DSA for assessment of the pedal vasculature, and is able to delineate significantly more patent vessels without segmental occlusions and more metatarsal arteries than selective DSA.     

Motion detection and correction using multi-rotation 180° single-photon emission tomography for thallium myocardial imaging

Patient and organ motion is a potentially limiting factor in gamma camera single-photon emission tomography (SPET) imaging, as highlighted in stress thallium myocardial SPET, where the heart may exhibit a systematic axial motion (cardiac creep) following stress. Multi-rotation SPET has previously been described as a means of obtaining better raw data for motion detection and correction. This study describes the validation of a computerised motion detection algorithm applied to multi-rotation SPET, and reports measured motions in thallium myocardial stress SPET studies from a single-headed gamma camera. Forty-two patients underwent pharmacological stress (dipyridamole) with leg raising, with injection of 75 MBq thallium-201 and imaging after a 10-min delay to detect or evaluate coronary artery disease. Multi-rotation gamma camera SPET was performed with a single-headed gamma camera, with five sequential rapid (4.5 min) continuous SPET mode rotations over 180°. A one-dimensional cross-correlation alignment technique was applied to the projection images to perform motion detection and correction in the axial direction prior to combining the five data sets for tomographic reconstruction. Validation of the cross-correlation alignment analysis was carried out by performing imaging with measured whole-body axial motions in nine subjects, and by reproducibility measurements on multi-rotation data sets. The effect of the applied motion correction was evaluated by calculating mean differences between image pairs before and after shifting, and the general reliability of the automatic motion detection was checked to within one pixel by visual assessment of 160 image pairs. Validation measurements of the cross-correlation technique gave a mean absolute error of 1.5±0.4 mm (0.24±0.06pixels) with a maximum error of 3.7 mm (0.6 pixels). In 40 subjects undergoing pharmacological stress ²⁰¹Tl myocardial SPET imaging, the mean cardiac axial creep movement was calculated as 3.1±0.7 mm (0.49±0.11 pixels), with 13 out of 40 (32%) having a calculated motion of 1 pixel (6.3 mm) or more. The automatic image shift was visually judged to be within 1 pixel in all 160 image pair analyses, and the mean pixel value difference between image pairs was reduced following image shifting. It is concluded that multi-rotation 180° SPET imaging provides raw data which allow objective and accurate motion detection of cardiac motion in thallium stress myocardial imaging, whilst the one-dimensional cross-correlation technique demonstrates adequate accuracy and reliability to be applied as an automatic motion screening technique on these data. Received 14 March and in revised form 27 July 1998     

DSA测量技术误差与控制

目的探讨DSA测量技术误差的产生与控制。方法在GE公司生产的Advantx LCV Plus DSA机上测得数据。将显示屏分为中央区、中间区及边远区3个区带，对不同大小定标在不同检查床高度及不同点光源与增强器高度（SID）条件下，测量标的（人民币5角硬币）的放大、缩小情况。结果随着标的外移，由中央区到边远区标的逐渐放大，且纵向放大比横向放大显著。不同的定标对比，硬币（直径20．4mm）和钢球（直径7．7mm）测量相同标的结果相差较小，而导管（4F）定标有显著的低估实物倾向。同区带同轴向测量，测量误差控制在1．0％～-2．5％之间。结论将显示屏划分为中央区、中间区及边远区将有助于介入医师对测量误差的控制。以定标物的横向做定标来测量标的较为准确，同区带同轴向测量误差控制较好。     

The AAPM/RSNA physics tutorial for residents: digital fluoroscopy.

A digital fluoroscopy system is most commonly configured as a conventional fluoroscopy system (tube, table, image intensifier, video system) in which the analog video signal is converted to and stored as digital data. Other methods of acquiring the digital data (eg, digital or charge-coupled device video and flat-panel detectors) will become more prevalent in the future. Fundamental concepts related to digital imaging in general include binary numbers, pixels, and gray levels. Digital image data allow the convenient use of several image processing techniques including last image hold, gray-scale processing, temporal frame averaging, and edge enhancement. Real-time subtraction of digital fluoroscopic images after injection of contrast material has led to widespread use of digital subtraction angiography (DSA). Additional image processing techniques used with DSA include road mapping, image fade, mask pixel shift, frame summation, and vessel size measurement. Peripheral angiography performed with an automatic moving table allows imaging of the peripheral vasculature with a single contrast material injection.     

Coronary calcium quantification using various calibration phantoms and scoring thresholds

RATIONALE AND OBJECTIVES: To compare scoring threshold and calibration method-dependent accuracy and variability of coronary calcium measurements by multidetector computed tomography (MDCT). METHODS: Ninety-five subjects were scanned with MDCT. We calculated Agatston score and volume score. Mineral mass (MM) was calculated using patient-based and scanner-based calibration methods. Accuracy of calibration was validated using artificial calcium cylinders. RESULTS: Patient-based and scanner-based calibration permitted accurate quantification of artificial calcium cylinders (bias: 0 mg and -2 mg). In the subjects, the mean relative difference of MM measurements performed at 90 and 130 Hounsfield units threshold (59%) was lower than for Agatston score (94%) and volume score (109%; P < 0.05). Patient-based and scanner-based calibration yielded systematically different MM measurements (bias: 22%). CONCLUSIONS: MM lowers threshold-dependent variability of coronary calcium measurements. Patient-based and scanner-based calibration allows accurate calcium quantification ex vivo but reveal systematic differences in subjects. Patient-based calibration may better account for subject size and composition.     

Method to correct for the effects of limited spatial resolution in phase-contrast flow MRI measurements.

Phase-contrast (PC) magnetic resonance imaging (MRI) flow measurements suffer from the effect of the point spread function (PSF) due to the limited sampling of k-space. The PSF, which in this case is a sinc function, deforms the flow profile and forms a ringing pattern around the vessel. In this work, an empirical method is presented that corrects for errors due to the deformation of the flow profile. The ringing pattern is used to obtain a well-defined vessel segmentation, which after correction provides more accurate vessel radius and volume flow rate (VFR). The correction method was developed from phantom measurements at constant flow and applied on phantom measurements at moderately pulsatile flow. After correction, the error of the estimated tube radius and the VFR was less than 10% and 5%, respectively. Corresponding errors without correction overestimated the radius by 60% and the VFR by 35%. Preliminary results indicate that the method is also valid in vivo. The variation in the estimated radius and VFR for different spatial resolution decreased when the method was applied. The presented method gives a more accurate estimation of the radius and VFR in vessels of the size of a few pixels without prior knowledge about the true vessel radius.     

Measurement of distensibility of blood vessels using cineangiograms

We developed a new, cineangiographic method to accurately measure the dynamic changes in the internal diameter of human arteries in vivo. Cine films were digitized at a spatial resolution of 4 microns/pixel, using a line image sensor. The vessel edges, with a Gaussian fit to a unilateral profile curve of the vessel, were determined with the aid of a computer program. We measured contrast-filled cylinder vessel models (2 to 7 mm in diameter) and evaluated precision, accuracy and linearity of the diameter measurement. A pulsatile vessel model of about 3.9 mm in internal diameter was used to examine the reliability of our method for detecting arterial wall motion. If the coefficient of variation of the vessel diameter determined cineangiographically was less than 2.2% we considered the cineangiograms sufficiently accurate to determine the internal vessel diameter and evaluate arterial distensibility.     

Scoutless stepping-table peripheral contrast-enhanced MR angiography

PURPOSE: To evaluate the feasibility of a scoutless method, termed EZ-STEP, for stepping-table peripheral contrast-enhanced magnetic resonance angiography (CE-MRA). MATERIALS AND METHODS: This scoutless method involves the use of a stepping-table, fast 3D MRA acquisition that incorporates spatially nonselective radiofrequency (RF) pulses for excitation to reduce the repetition time (TR). The sequence was tested in a phantom. The EZ-STEP protocol was optimized in four healthy volunteers and used in 15 subjects. The image quality was scored in a blinded fashion and compared with conventional MRA in eight patients. RESULTS: The acquisition speed of the EZ-STEP sequence was approximately 30% faster in the phantom study compared to the conventional MRA sequence. The total examination time for EZ-STEP was 6 minutes, compared to an average of 23 minutes for conventional MRA. The average image quality scores for EZ-STEP and conventional MRA for stations 1-3 were 3.50 vs. 3.06 (P = 0.087), 3.53 vs. 3.00 (P = 0.033), and 2.97 vs. 2.50 (P = 0.090), respectively. CONCLUSION: EZ-STEP is a more efficient method than the conventional approach for stepping-table peripheral CE-MRA, and provides comparable or better image quality. This method shortens the examination time substantially and eliminates the risk of failing to image a vessel because of improper positioning of the scan volume.     

Prospective motion correction with continuous gradient updates in diffusion weighted imaging

Despite the existence of numerous motion correction methods, head motion during MRI continues to be a major source of artifacts and can greatly reduce image quality. This applies particularly to diffusion weighted imaging, where strong gradients are applied during long encoding periods. These are necessary to encode microscopic movements. However, they also make the technique highly sensitive to bulk motion. In this work, we present a prospective motion correction method where all applied gradients are adjusted continuously to compensate for changes of the object position and ensure the desired phase evolution in the image coordinate frame. Additionally, in phantom experiments this new technique is used to reproduce motion artifacts with high accuracy by changing the position of the imaging frame relative to the measured object. In vivo measurements demonstrate the validity of the new correction method.     

Left ventricular volume determined from scintigraphy and digital angiography by a semi-automated geometric method

The utility of a semi-automatic method of measuring left ventricular (LV) volume geometrically from gated blood-pool studies and digital subtraction angiography (DSA) was investigated using computerized edge detection and spatial calibration algorithms. LAO LV volumes determined from gated blood-pool studies were compared to volumes obtained from contrast left ventriculograms in 21 patients and the applicability of this method to DSA was evaluated in 25 additional patients who also had conventional left ventriculography. There was excellent correlation between the two, both for radionuclide studies and for DSA. Computer-based geometric determinations of LV volume appear to be rapid, accurate, and less dependent on subjective operator decisions than previously reported geometric approaches.     

自发性蛛网膜下腔出血的CTA与DSA对比研究

目的:探讨64层CTA与DSA对自发性蛛网膜下腔出血(spontaneous subarachnoid hemorrhage,s-SAH)的应用价值及其限度。方法:收集153例s-SAH患者的CTA及DSA检查资料,回顾性分析2种检查方法的图像质量(采取1³分评分制),并比较2种方法在脑动脉瘤检出方面的灵敏度、特异度及准确度。结果:图像质量评分CTA为(2.63±0.262)分,DSA为(2.73±0.254)分,二者差异无统计学意义(P>0.05)。CTA对脑动脉瘤的检出的灵敏度为96.2%,特异度为94.9%,准确度为95.9%;DSA灵敏度为97.7%,特异度为97.4%,准确度为97.6%,二者差异无统计学意义(P>0.05)。结论:在s-SAH患者中,CTA与DSA均能准确检出动脉瘤所致破裂出血。相较于DSA,CTA能多方位显示病变,对血管壁及血管周围情况的显示更具价值。     

A 3D space-time motion evaluation for image registration in digital subtraction angiography.

In modern clinical practice, Digital Subtraction Angiography (DSA) is a powerful technique for the visualization of blood vessels in a sequence of X-ray images. A serious problem encountered in this technique is the presence of artifacts due to patient motion. The resulting artifacts frequently lead to misdiagnosis or rejection of a DSA image sequence. In this paper, a new technique for removing both global and local motion artifacts is presented. It is based on a 3D space-time motion evaluation for separating pixels changing values because of motion from those changing values because of contrast flow. This technique is proved to be very efficient to correct for patient motion artifacts and is computationally cheap. Experimental results with several clinical data sets show that this technique is very fast and results in higher quality images.     

Effect of acquisition parameters on the accuracy of velocity encoded cine magnetic resonance imaging blood flow measurements.

PURPOSE: To investigate the effect of acquisition parameters on the accuracy of 2D velocity encoded cine magnetic resonance imaging (VEC MRI) flow measurements. MATERIALS AND METHODS: Using a pulsatile flow phantom, through-plane flow measurements were performed on a flexible vessel made of polyvinyl alcohol cryogel (PVA), a material that mimics the MR signal and biomechanical properties of aortic tissue. RESULTS: Repeated VEC MRI flow measurements (N = 20) under baseline conditions yielded an error of 0.8 +/- 1.5%. Slice thickness, angle between flow and velocity encoding directions, spatial resolution, velocity encoding range, and radio frequency (RF) flip angles were varied over a clinically relevant range. Spatial resolution had the greatest impact on accuracy, with a 9% overestimation of flow at 16 pixels per vessel cross-section. CONCLUSION: VEC MRI proved to be an accurate and reproducible technique for pulsatile flow measurements over the range of acquisition parameters examined as long as sufficient spatial resolution was prescribed.     

Multi-scale retinal vessel segmentation using line tracking

In this paper an algorithm for vessel segmentation and network extraction in retinal images is proposed. A new multi-scale line-tracking procedure is starting from a small group of pixels, derived from a brightness selection rule, and terminates when a cross-sectional profile condition becomes invalid. The multi-scale image map is derived after combining the individual image maps along scales, containing the pixels confidence to belong in a vessel. The initial vessel network is derived after map quantization of the multi-scale confidence matrix. Median filtering is applied in the initial vessel network, restoring disconnected vessel lines and eliminating noisy lines. Finally, post-processing removes erroneous areas using directional attributes of vessels and morphological reconstruction.The experimental evaluation in the publicly available DRIVE database shows accurate extraction of vessels network. The average accuracy of 0.929 with 0.747 sensitivity and 0.955 specificity is very close to the manual segmentation rates obtained by the second observer. The proposed algorithm is compared also with widely used supervised and unsupervised methods and evaluated in noisy conditions, giving higher average sensitivity rate in the same range of specificity and accuracy, and showing robustness in the presence of additive Salt&Pepper or Gaussian white noise.     

Quantitative digital subtraction arteriography with a calibration catheter

Quantitative intraarterial digital subtraction arteriography (DSA) was performed using a calibration catheter with three distal metallic ring markers. The two outer markers were 50 mm apart, and the third marker was in the middle. Measurements of 54 vessel diameters of the abdominal aorta, renal, lumbar, and iliac arteries were performed in a comparison study with direct film arteriograms in 10 aortofemoral runoff studies. Diameter measurements were made by both the observer on hard copy DSA images and by a computer using modified semiquantitative software. Against measurements on film, which were used as the standard, deviations in measurement on digital images varied from 8 to 13% for arterial diameters <5 mm and from 2 to 6% for diameters ≥5 mm. Projectional artifacts caused 3% or less error. Knowing these variations in measurement is important in order to determine error tolerances for clinical applications. The calibration catheter serves as a convenient internal marker for DSA.     

One- and two-dimensional EPR imaging studies on phantoms and plant specimens

The advantages of electron paramagnetic resonance (EPR) imaging at L-band frequencies are discussed. The construction and calibration of a low-field L-band EPR imaging spectrometer is described with capillary phantoms containing aqueous nitroxides as paramagnetic imaging agents. The peak separation induced by the magnetic field gradient is related to the object separation. The linewidth of each (first-derivative) line is used to calculate the dimensions of each paramagnetic object, which is the sum of an intrinsic and a gradient-induced component. By extrapolating the linewidth back to zero field gradient, one obtains the intrinsic linewidth correction factor in computing image size. The nonuniformity caused by deterioration of biological substructure was examined with plant stems where one capillary vessel had become leaky. Spin-label destruction, particularly by biological reducing agents, was compared for three species of plant stems.