Monoplane 3D-2D registration of cerebral angiograms based on multi-objective stratified optimization.

Registration of 3D pre-interventional to 2D intra-interventional medical images has an increasingly important role in surgical planning, navigation and treatment, because it enables the physician to co-locate depth information given by pre-interventional 3D images with the live information in intra-interventional 2D images such as x-ray. Most tasks during image-guided interventions are carried out under a monoplane x-ray, which is a highly ill-posed problem for state-of-the-art 3D to 2D registration methods. To address the problem of rigid 3D-2D monoplane registration we propose a novel multi-objective stratified parameter optimization, wherein a small set of high-magnitude intensity gradients are matched between the 3D and 2D images. The stratified parameter optimization matches rotation templates to depth templates, first sampled from projected 3D gradients and second from the 2D image gradients, so as to recover 3D rigid-body rotations and out-of-plane translation. The objective for matching was the gradient magnitude correlation coefficient, which is invariant to in-plane translation. The in-plane translations are then found by locating the maximum of the gradient phase correlation between the best matching pair of rotation and depth templates. On twenty pairs of 3D and 2D images of ten patients undergoing cerebral endovascular image-guided intervention the 3D to monoplane 2D registration experiments were setup with a rather high range of initial mean target registration error from 0 to 100 mm. The proposed method effectively reduced the registration error to below 2 mm, which was further refined by a fast iterative method and resulted in a high final registration accuracy (0.40 mm) and high success rate ([Formula: see text]96%). Taking into account a fast execution time below 10 s, the observed performance of the proposed method shows a high potential for application into clinical image-guidance systems.

Manual and computerized measurement of coronal vertebral inclination on MRI images: a pilot study.

AIM: A pilot study that presents a systematic approach for evaluating the variability of manual and computerized measurements of coronal vertebral inclination (CVI) on images acquired by magnetic resonance imaging (MRI). MATERIALS AND METHODS: Three observers identified the vertebral body corners of 28 vertebrae on two occasions on two-dimensional (2D) coronal MRI cross-sections, which served to evaluate CVI using six manual measurements (superior and inferior tangents, left and right tangents, mid-endplate and mid-wall lines). Computerized measurements were performed by evaluating CVI from the symmetry of vertebral anatomical structures of the same 28 vertebrae in 2D coronal MRI cross-sections and in three-dimensional (3D) MRI images. RESULTS: In terms of standard deviation (SD), the mid-endplate lines proved to be the manual measurements with the lowest intra- (1.0° SD) and interobserver (1.4° SD) variability. The computerized measurements in 3D yielded even lower intra- (0.8° SD) and interobserver (1.3° SD) variability. The strongest inter-method agreement (1.2° SD) was found among lines parallel to vertebral endplates (superior tangents, inferior tangents, mid-endplate lines). The computerized measurements in 3D were most in agreement with the mid-endplate lines (1.9° SD). The estimated intra- and interobserver variabilities of standard Cobb angle measurements were equal to 1.6° SD and 2.5° SD, respectively, for manual measurements, and to 1.1° SD and 1.8° SD, respectively, for computerized measurements. CONCLUSION: The mid-endplate lines proved to be the most reproducible and reliable manual CVI measurements. Computerized CVI measurements based on the evaluation of the symmetry of vertebral anatomical structures in 3D were more reproducible and reliable than manual measurements.

Quantitative vertebral morphometry based on parametric modeling of vertebral bodies in 3D.

UNLABELLED: Quantitative vertebral morphometry (QVM) was performed by parametric modeling of vertebral bodies in three dimensions (3D). INTRODUCTION: Identification of vertebral fractures in two dimensions is a challenging task due to the projective nature of radiographic images and variability in the vertebral shape. By generating detailed 3D anatomical images, computed tomography (CT) enables accurate measurement of vertebral deformations and fractures. METHODS: A detailed 3D representation of the vertebral body shape is obtained by automatically aligning a parametric 3D model to vertebral bodies in CT images. The parameters of the 3D model describe clinically meaningful morphometric vertebral body features, and QVM in 3D is performed by comparing the parameters to their statistical values. Thresholds and parameters that best discriminate between normal and fractured vertebral bodies are determined by applying statistical classification analysis. RESULTS: The proposed QVM in 3D was applied to 454 normal and 228 fractured vertebral bodies, yielding classification sensitivity of 92.5% at 7.5% specificity, with corresponding accuracy of 92.5% and precision of 86.1%. The 3D shape parameters that provided the best separation between normal and fractured vertebral bodies were the vertebral body height and the inclination and concavity of both vertebral endplates. CONCLUSION: The described QVM in 3D is able to efficiently and objectively discriminate between normal and fractured vertebral bodies and identify morphological cases (wedge, (bi)concavity, or crush) and grades (1, 2, or 3) of vertebral body fractures. It may be therefore valuable for diagnosing and predicting vertebral fractures in patients who are at risk of osteoporosis.

A review of 3D/2D registration methods for image-guided interventions.

Registration of pre- and intra-interventional data is one of the key technologies for image-guided radiation therapy, radiosurgery, minimally invasive surgery, endoscopy, and interventional radiology. In this paper, we survey those 3D/2D data registration methods that utilize 3D computer tomography or magnetic resonance images as the pre-interventional data and 2D X-ray projection images as the intra-interventional data. The 3D/2D registration methods are reviewed with respect to image modality, image dimensionality, registration basis, geometric transformation, user interaction, optimization procedure, subject, and object of registration.

Validation for 2D/3D registration. I: A new gold standard data set.

PURPOSE: In this article, the authors propose a new gold standard data set for the validation of two-dimensional/three-dimensional (2D/3D) and 3D/3D image registration algorithms. METHODS: A gold standard data set was produced using a fresh cadaver pig head with attached fiducial markers. The authors used several imaging modalities common in diagnostic imaging or radiotherapy, which include 64-slice computed tomography (CT), magnetic resonance imaging using T1, T2, and proton density sequences, and cone beam CT imaging data. Radiographic data were acquired using kilovoltage and megavoltage imaging techniques. The image information reflects both anatomy and reliable fiducial marker information and improves over existing data sets by the level of anatomical detail, image data quality, and soft-tissue content. The markers on the 3D and 2D image data were segmented using ANALYZE 10.0 (AnalyzeDirect, Inc., Kansas City, KN) and an in-house software. RESULTS: The projection distance errors and the expected target registration errors over all the image data sets were found to be less than 2.71 and 1.88 mm, respectively. CONCLUSIONS: The gold standard data set, obtained with state-of-the-art imaging technology, has the potential to improve the validation of 2D/3D and 3D/3D registration algorithms for image guided therapy.

Efficient implementation of the rank correlation merit function for 2D/3D registration.

A growing number of clinical applications using 2D/3D registration have been presented recently. Usually, a digitally reconstructed radiograph is compared iteratively to an x-ray image of the known projection geometry until a match is achieved, thus providing six degrees of freedom of rigid motion which can be used for patient setup in image-guided radiation therapy or computer-assisted interventions. Recently, stochastic rank correlation, a merit function based on Spearman's rank correlation coefficient, was presented as a merit function especially suitable for 2D/3D registration. The advantage of this measure is its robustness against variations in image histogram content and its wide convergence range. The considerable computational expense of computing an ordered rank list is avoided here by comparing randomly chosen subsets of the DRR and reference x-ray. In this work, we show that it is possible to omit the sorting step and to compute the rank correlation coefficient of the full image content as fast as conventional merit functions. Our evaluation of a well-calibrated cadaver phantom also confirms that rank correlation-type merit functions give the most accurate results if large differences in the histogram content for the DRR and the x-ray image are present.

Comparative evaluation of retrospective shading correction methods.

Because of the inherent imperfections of the image formation process, microscopical images are often corrupted by spurious intensity variations. This phenomenon, known as shading or intensity inhomogeneity, may have an adverse affect on automatic image processing, such as segmentation and registration. Shading correction methods may be prospective or retrospective. The former require an acquisition protocol tuned to shading correction, whereas the latter can be applied to any image, because they only use the information already present in an image. Nine retrospective shading correction methods were implemented, evaluated and compared on three sets of differently structured synthetic shaded and shading-free images and on three sets of real microscopical images acquired by different acquisition set-ups. The performance of a method was expressed quantitatively by the coefficient of joint variations between two different object classes. The results show that all methods, except the entropy minimization method, work well for certain images, but perform poorly for others. The entropy minimization method outperforms the other methods in terms of reduction of true intensity variations and preservation of intensity characteristics of shading-free images. The strength of the entropy minimization method is especially apparent when applied to images containing large-scale objects.

Comparative evaluation of JPEG and JPEG2000 compression in quantitative digital subtraction radiography.

OBJECTIVES: The aim of this in vitro study was to compare the impact of JPEG and the novel JPEG2000 compression standard on quantitative digital subtraction radiography (DSR) and to determine the acceptable JPEG2000 compression ratios for DSR. METHODS: Nine dry pig mandible sections were radiographed three times ('Baseline', 'No change', and 'Gain') with standardized projection geometry. Bone gain was simulated by adding artificial bone chips (1, 4 and 15 mg). Images were registered, compressed by JPEG and JPEG2000 using compression ratios (CR) of 1 : 7, 1 : 16, 1 : 22, and 1 : 31, and then subtracted. Image distortion was assessed objectively by calculating average pixel error and peak signal to noise ratio. No change areas in compressed and subtracted 'No change-Baseline' images and bone gain volumes in compressed and subtracted 'Gain-Baseline' images were calculated for both compression standards and compared. RESULTS: JPEG introduced less distortion at low CRs, while JPEG2000 was superior at higher CRs. At CR of 1 : 7, no significant difference between JPEG and JPEG2000 was found. JPEG2000 yielded better results for no change measurements at higher CRs. Volumes of simulated bone gain were overestimated when JPEG and underestimated when JPEG2000 compression was used. CONCLUSIONS: At CR of 1 : 7 JPEG and JPEG2000 performed similarly, which indicates that CR of 1:7 in JPEG2000 can be used for DSR if images are registered before compression. At higher CRs, JPEG2000 is superior to JPEG but image distortions are too high for reliable quantitative DSR.

Impact of JPEG lossy image compression on quantitative digital subtraction radiography.

OBJECTIVES: The aim of the study was to evaluate the impact of JPEG lossy image compression on the estimation of alveolar bone gain by quantitative digital subtraction radiography (DSR). METHODS: Nine dry domestic pig mandible posterior segments were radiographed three times ('Baseline', 'No change', and 'Gain') with standardized projection geometry. Bone gain was simulated by adding artificial bone chips (1, 4, and 15 mg). Images were either compressed before or after registration. No change areas in compressed and subtracted 'No change-Baseline' images and bone gain volumes in compressed and subtracted 'Gain-Baseline' images were calculated and compared to the corresponding measurements performed on original subtracted images. RESULTS: Measurements of no change areas ('No change-Baseline') were only slightly affected by compressions down to JPEG 50 (J50) applied either before or after registration. Simulated gain of alveolar bone ('Gain-Baseline') was underestimated when compression before registration was performed. The underestimation was bigger when small bone chips of 1 mg were measured and when higher compression rates were used. Bone chips of 4 and 15 mg were only slightly underestimated when using J90, J70, and J50 compressions before registration. CONCLUSIONS: Lossy JPEG compression does not affect the measurements of no change areas by DSR. Images undergoing subtraction should be registered before compression and if so, J90 compression with a compression ratio of 1:7 can be used to detect and measure 4 mg and larger bone gain.

Model-based automated detection of mammalian cell colonies.

Manually counting cell colonies, especially those that originate from fibroblast cell lines, is a time-consuming, eye-straining and tedious task in which consistency of counting is difficult to maintain. In this paper we present a novel model-based image segmentation method, which employs prior knowledge about the shape of a colony with the aim to automatically detect isolated, touching and overlapping cell colonies of various sizes and intensities. First, a set of hypothetical model instances is generated by using a robust statistical approach to estimate the model parameters and a novel confidence measure to quantify the difference between a model instance and the underlying image. Second, the model instances matching the individual colonies in the image are selected from the set by a minimum description length principle. The procedure was applied to images of Chinese hamster lung fibroblast cell line DC3F, which forms poorly defined or 'fuzzy' colonies. The correlation with manual counting was determined and the cell survival curves obtained by automated and manual counting were compared. The results obtained show that the proposed automatic procedure was capable to correctly identify 91% of cell colonies typical of mammalian cell lines.

Retrospective correction of MR intensity inhomogeneity by information minimization.

In this paper, the problem of retrospective correction of intensity inhomogeneity in magnetic resonance (MR) images is addressed. A novel model-based correction method is proposed, based on the assumption that an image corrupted by intensity inhomogeneity contains more information than the corresponding uncorrupted image. The image degradation process is described by a linear model, consisting of a multiplicative and an additive component which are modeled by a combination of smoothly varying basis functions. The degraded image is corrected by the inverse of the image degradation model. The parameters of this model are optimized such that the information of the corrected image is minimized while the global intensity statistic is preserved. The method was quantitatively evaluated and compared to other methods on a number of simulated and real MR images and proved to be effective, reliable, and computationally attractive. The method can be widely applied to different types of MR images because it solely uses the information that is naturally present in an image, without making assumptions on its spatial and intensity distribution. Besides, the method requires no preprocessing, parameter setting, nor user interaction. Consequently, the proposed method may be a valuable tool in MR image analysis.

Influence of developer exhaustion on accuracy of quantitative digital subtraction radiography: an in vitro study.

OBJECTIVE: The aim of the study was to evaluate the influence of developer exhaustion on accuracy of quantitative digital subtraction radiography. STUDY DESIGN: Six objects, each incorporating a section of dry human mandible, were radiographed with 4 exposure times. Baseline films were processed in fresh solutions, whereas follow-up films were processed in fresh and in increasingly exhausted solutions (ie, 1, 2, and 3 weeks old). Bone loss and bone gain were computer simulated in 17 regions of interest on baseline radiographs. Area and volume of changes in mineralization were measured in subtracted images, obtained by subtraction of baseline from their corresponding follow-up radiographs. Friedman's 2-way analysis of variance by ranks and Wilcoxon signed-rank test were used for statistical analysis. RESULTS: Because of exhausted developer, bone loss was relatively underestimated from 6.6% to 16.5% (P <.05), whereas bone gain was relatively overestimated from 9.7% to 16.7% (P <.05). CONCLUSIONS: This in vitro study demonstrates that films for quantitative digital subtraction radiography should be processed in fresh developer or error might be introduced.

Retrospective shading correction based on entropy minimization.

Shading is a prominent phenomenon in microscopy, manifesting itself via spurious intensity variations not present in the original scene. The elimination of shading effects is frequently necessary for subsequent image processing tasks, especially if quantitative analysis is the final goal. While most of the shading effects may be minimized by setting up the image acquisition conditions carefully and capturing additional calibration images, object-dependent shading calls for retrospective correction. In this paper a novel method for retrospective shading correction is proposed. Firstly, the image formation process and the corresponding shading effects are described by a linear image formation model, which consists of an additive and a multiplicative parametric component. Secondly, shading correction is performed by the inverse of the image formation model, whose shading components are estimated retrospectively by minimizing the entropy of the acquired images. A number of tests, performed on artificial and real microscopical images, show that this approach is efficient for a variety of differently structured images and as such may have applications in and beyond the field of microscopical imaging.

Automatic extraction of corresponding points for the registration of medical images.

In this paper we address the problem of finding corresponding points in a reference and its subsequent image with the aim of registering the images. A whole-image-content-based automatic algorithm for extracting point pairs from 2-D monomodal medical images has been developed. The properties of point distinctiveness, point pair similarity, and point pair consistency have been incorporated into the steps which lead to the automatic extraction and weighting of point pairs. The selection of the most distinctive points of the reference image, and the search for their corresponding points in the subsequent image, have two things in common. First, the local operator by which the distinctive points are selected mimics the template matching used to find the corresponding points. Second, the same similarity measure is used for both tasks. We have applied the algorithm to a variety of computer-generated and real medical images, and have both qualitatively and quantitatively evaluated its performance. The results show that the proposed automatic algorithm for point extraction is accurate and robust and that it may significantly improve on the accuracy, reproducibility, and speed of the manual extraction of corresponding points.

Registration of serial transverse sections of muscle fibers.

BACKGROUND: Structural and functional characterization of skeletal muscles is often assessed by histochemical techniques, which enable the classification into different fiber types by combining the reactions in serial transverse muscle cross-sections. A drawback is that a knowledgeable operator is required to combine and evaluate the reactions, which is a time-consuming, tedious, and subjective task. To enhance the speed and reproducibility of muscle fiber typing, the registration of serial transverse sections of muscle fibers images has been proposed as a preprocessing step. METHODS: Three different registration methods were considered: first, a semi-automatic elastic point-based registration; second, an automatic (rigid, affine, and projective) whole image content-based registration; and, third, an automatic hierarchical elastic registration obtained by integration of the first two methods. The performances of the methods were tested on a database of 50 image stacks each containing three images of histochemically differently stained serial human muscle cross-sections. RESULTS: The amounts of successful globally and locally registered stacks were approximately 20% for automatic rigid registration, 60% for automatic affine and projective registrations, and 80% for semi-automatic and automatic elastic registration. CONCLUSIONS: By using robust elastic registration methods, the automatic registration of serial transverse muscle fiber images seems feasible and might allow automatic muscle fiber typing, and consequently, improve the characterization of skeletal muscles.

Automated segmentation of muscle fiber images using active contour models.

The cross-sectional area of different fiber types is an important anatomic feature in studying the structure and function of healthy and diseased human skeletal muscles. However, such studies are hampered by the thousands of fibers involved when manual segmentation has to be used. We have developed a semiautomatic segmentation method that uses computational geometry and recent computer vision techniques to significantly reduce the time required to accurately segment the fibers in a sample. The segmentation is achieved by simply pointing to the approximate centroid of each fiber. The set of centroids is then used to automatically construct the Voronoi polygons, which correspond to individual fibers. Each Voronoi polygon represents the initial shape of one active contour model, called a snake. In the energy minimization process, which is executed in several stages, different external forces and problem-specific knowledge are used to guide the snakes to converge to fiber boundaries. Our results indicate that this approach for segmenting muscle fiber images is fast, accurate, and reproducible compared with manual segmentation performed by experts.

Evaluation of three contrast correction methods for digital subtraction in dental radiography: an in vitro study.

In the present in vitro study we test the efficacy of three most often used contrast correction methods to reduce contrast mismatches that can adversely affect digital subtraction in dental radiography. Five radiographs of the lower molars of a beagle dog with an included reference aluminium wedge were radiographed with different exposure times, captured by a monochrome CCD video camera and digitized. To study only the contrast correction effects, the influences of projective geometry and physical noise from the camera and the analog-to-digital conversion were eliminated by simulating bone loss in digital copies of the five reference radiographic images. Each copy in which loss was simulated was contrast corrected by all methods and subtracted from its original image. Four parameters were defined to test the potential of a method to correct grey level values between two images without corrupting real changes. The ODTF (optical density thickness function) method, which is based on a function relating grey level values of the aluminium wedge image and the corresponding thicknesses of the wedge, induced less contrast correction error than the CDF (cumulative density function) and the LSQA (least square quadratic approximation) methods. Therefore, the ODTF method may detect more subtle changes than the other two methods. The experiments indicate that less error is induced by a contrast correction method which uses a grey level mapping function based on the density distribution in a region representing a reference structure. Moreover, CDF, ODTF, and LSQA functions obtained from the reference structure density distribution may be applied for objective contrast enhancements and for standardization of image quality, while the ODTF function also allows bone change volume estimations.

Determination of the femoral and pelvic geometrical parameters that are important for the hip joint contact stress: differences between female and male.

The difference between male and female femoral and pelvic geometry was studied by considering some geometrical parameters such as interhip distance, inclination of the crista iliaca, diameter of the femoral head and centre-edge angle of Wiberg. The values of these parameters were determined for 79 healthy female and 21 healthy male subjects. Standard anterior-posterior radiographs were used and processed by the computer-aided system. The results show some important sex differences in femoral and pelvic geometrical parameters which determine the hip joint contact stress.

Energy metabolism of fibre types within fascicles of human muscles.

In human latissimus dorsi muscle a preponderance of type 2b fibres in the first fascicle layer and of type 1 fibres in the second layer was found. NADH-dehydrogenase (NADH) and alpha-glycerophosphate dehydrogenase (GPDH) which were measured histophotometrically in type 1, 2a, and 2b fibres showed either extreme or only partial overlapping regarding the activity of metabolic enzymes. In different fascicle layers the average activity of both enzymes did not differ significantly among the fibres of the same type, neither did the NADH and GPDH activity of type 2a and 2b.

Contrast matching techniques for digital subtraction radiography: an objective evaluation.

Digital subtraction radiography (DSR) enables the detection of subtle early detrimental effects of periodontal disease as well as the evaluation of the effects of therapy. However, the differences between two radiographs due to alignment and contrast errors must be kept at minimum. In the present in vitro study we test the efficacy of three basic contrast correction methods in the reduction of contrast mismatches which can adversely affect a subtracted image. The ODTF (Optical Density Thickness Function) method, which is based on a function relating grey level values of the aluminium wedge image and the corresponding thickness of the wedge, induced less contrast correction error than the CDF (Cumulative Density Function) and the LSQA (Least Square Quadratic Approximation) methods. Moreover, CDF, ODTF, and LSQA functions obtained from the reference structure density distribution may be applied for objective contrast enhancements and for standardisation of image quality, while the ODTF function allows also bone change volume estimations.