Accurate assessment of patellar tracking using fiducial and intensity-based fluoroscopic techniques

Fully automatic tracking of native glenohumeral kinematics from stereo-radiography

The current work introduces a system for fully automatic tracking of native glenohumeral kinematics in stereo-radiography sequences. The proposed method first applies convolutional neural networks to obtain segmentation and semantic key point predictions in biplanar radiograph frames. Preliminary bone pose estimates are computed by solving a non-convex optimization problem with semidefinite relaxations to register digitized bone landmarks to semantic key points. Initial poses are then refined by registering computed tomography-based digitally reconstructed radiographs to captured scenes, which are masked by segmentation maps to isolate the shoulder joint. A particular neural net architecture which exploits subject-specific geometry is also introduced to improve segmentation predictions and increase robustness of subsequent pose estimates. The method is evaluated by comparing predicted glenohumeral kinematics to manually tracked values from 17 trials capturing 4 dynamic activities. Median orientation differences between predicted and ground truth poses were 1.7∘ and 8.6∘ for the scapula and humerus, respectively. Joint-level kinematics differences were less than 2∘ in 65%, 13%, and 63% of frames for XYZ orientation DoFs based on Euler angle decompositions. Automation of kinematic tracking can increase scalability of tracking workflows in research, clinical, or surgical applications.

A new 2D-3D registration gold-standard dataset for the hip joint based on uncertainty modeling

Purpose

Estimation of the accuracy of 2D‐3D registration is paramount for a correct evaluation of its outcome in both research and clinical studies. Publicly available datasets with standardized evaluation methodology are necessary for validation and comparison of 2D‐3D registration techniques. Given the large use of 2D‐3D registration in biomechanics, we introduced the first gold standard validation dataset for computed tomography (CT)‐to‐x‐ray registration of the hip joint, based on fluoroscopic images with large rotation angles. As the ground truth computed with fiducial markers is affected by localization errors in the image datasets, we proposed a new methodology based on uncertainty propagation to estimate the accuracy of a gold standard dataset.

Methods

The gold standard dataset included a 3D CT scan of a female hip phantom and 19 2D fluoroscopic images acquired at different views and voltages. The ground truth transformations were estimated based on the corresponding pairs of extracted 2D and 3D fiducial locations. These were assumed to be corrupted by Gaussian noise, without any restrictions of isotropy. We devised the multiple projective points criterion (MPPC) that jointly optimizes the transformations and the noisy 3D fiducial locations for all views. The accuracy of the transformations obtained with the MPPC was assessed in both synthetic and real experiments using different formulations of the target registration error (TRE), including a novel formulation of the TRE (uTRE) derived from the uncertainty analysis of the MPPC.

Results

The proposed MPPC method was statistically more accurate compared to the validation methods for 2D‐3D registration that did not optimize the 3D fiducial positions or wrongly assumed the isotropy of the noise. The reported results were comparable to previous published works of gold standard datasets. However, a formulation of the TRE commonly found in these gold standard datasets was found to significantly miscalculate the true TRE computed in synthetic experiments with known ground truths. In contrast, the uncertainty‐based uTRE was statistically closer to the true TRE.

Conclusions

We proposed a new gold standard dataset for the validation of CT‐to‐X‐ray registration of the hip joint. The gold standard transformations were derived from a novel method modeling the uncertainty in extracted 2D and 3D fiducials. Results showed that considering possible noise anisotropy and including corrupted 3D fiducials in the optimization resulted in improved accuracy of the gold standard. A new uncertainty‐based formulation of the TRE also appeared as a good alternative to the unknown true TRE that has been replaced in previous works by an alternative TRE not fully reflecting the gold standard accuracy.

Can Computer-Assisted Total Knee Arthroplasty Support the Prediction of Postoperative Three-Dimensional Kinematics of the Tibiofemoral and Patellofemoral Joints at the Replaced Knee?

The aim of this study was to analyze the extent to which postoperative patellofemoral joint (PFJ) kinematics assessed at 6-month follow-up after total knee arthroplasty (TKA) mimics the intraoperative kinematics after final component implantation. The study hypothesis, already proved in terms of tibiofemoral joint (TFJ) kinematics, is that the intraoperative assessment of PFJ kinematics after component implantation is also capable of predicting postoperative knee kinematics during activities of daily living. Twenty patients selected for TKA with patellar resurfacing were implanted using surgical navigation, including patellar component positioning via a novel computer-assisted procedure. This allowed for intraoperative TFJ and PFJ kinematic assessment after final component implantation. At 6-month follow-up, all patients were contacted for follow-up control; in addition to clinical examination, this implied postoperative kinematics assessments by three-dimensional video fluoroscopy of the replaced knee during standard activities of daily living. Several traditional PFJ, as well as TFJ, rotations and translations were calculated intra- and postoperatively and then statistically compared. Good postoperative replication of the intraoperative measurements was observed for most of PFJ variables analyzed, as well as those for TFJ. Relevant statistical analysis also supported the significant consistency between the intra- and postoperative measurements. Pertaining to the present findings on a statistical basis, intraoperative measurements performed at both TFJ and PFJ kinematics using a surgical navigation system under passive conditions, are predictive of the overall knee kinematics experienced at postoperative follow-ups by the same replaced knees in typical activities of daily living.

Fast and automatic periacetabular osteotomy fragment pose estimation using intraoperatively implanted fiducials and single-view fluoroscopy

Accurate and consistent mental interpretation of fluoroscopy to determine the position and orientation of acetabular bone fragments in 3D space is difficult. We propose a computer assisted approach that uses a single fluoroscopic view and quickly reports the pose of an acetabular fragment without any user input or initialization. Intraoperatively, but prior to any osteotomies, two constellations of metallic ball-bearings (BBs) are injected into the wing of a patient's ilium and lateral superior pubic ramus. One constellation is located on the expected acetabular fragment, and the other is located on the remaining, larger, pelvis fragment. The 3D locations of each BB are reconstructed using three fluoroscopic views and 2D/3D registrations to a preoperative CT scan of the pelvis. The relative pose of the fragment is established by estimating the movement of the two BB constellations using a single fluoroscopic view taken after osteotomy and fragment relocation. BB detection and inter-view correspondences are automatically computed throughout the processing pipeline. The proposed method was evaluated on a multitude of fluoroscopic images collected from six cadaveric surgeries performed bilaterally on three specimens. Mean fragment rotation error was 2.4 ± 1.0 degrees, mean translation error was 2.1 ± 0.6 mm, and mean 3D lateral center edge angle error was 1.0 ± 0.5 degrees. The average runtime of the single-view pose estimation was 0.7 ± 0.2 s. The proposed method demonstrates accuracy similar to other state of the art systems which require optical tracking systems or multiple-view 2D/3D registrations with manual input. The errors reported on fragment poses and lateral center edge angles are within the margins required for accurate intraoperative evaluation of femoral head coverage.

Augmented reality and artificial intelligence-assisted surgical navigation: Technique and cadaveric feasibility study

Purpose

Augmented reality-based image overlay of virtual bony spine anatomy can be projected onto real spinal anatomy using computer tomography-generated DICOM images acquired intraoperatively. The aim of the study was to develop a technique and assess the accuracy and feasibility of lumbar vertebrae pedicle instrumentation using augmented reality-assisted surgical navigation.

Subjects and Methods

An augmented reality and artificial intelligence (ARAI)-assisted surgical navigation system was developed. The system consists of a display system which hovers over the surgical field and projects three-dimensional (3D) medical images corresponding with the patient's anatomy. The system was registered to the cadaveric spine using an optical tracker and arrays with reflective markers. The virtual image overlay from the ARAI system was compared to 3D generated images from intraoperative scans and used to percutaneously navigate a probe to the cortex at the corresponding pedicle starting point. Intraoperative scan was used to confirm the probe position. Virtual probe placement was compared to the actual probe position in the bone to determine the accuracy of the navigation system.

Results

Four cadaveric thoracolumbar spines were used. The navigated probes were correctly placed in all attempted levels (n = 24 levels), defined as Zdichavsky type 1a, Ravi type I, and Gertzbein type 0. The virtual overlay image corresponded to the 3D generated image in all the tested levels.

Conclusions

The ARAI surgical navigation system correctly and accurately identified the starting points at all the attempted levels. The virtual anatomy image overlay precisely corresponded to the actual anatomy in all the tested scenarios. This technology may lead more uniform outcomes between surgeons and decrease minimally invasive spine surgery learning curves.

Development and validation of a robust patellar reference coordinate system for biomechanical and clinical studies

BACKGROUND

This study aimed to develop and validate a reference coordinate system for the human patella, based on the registration of bony landmarks on a computed tomography (CT) scan.

METHODS

Thirty-three native cadaveric specimens were scanned, and an observer marked a set of seven anatomical landmarks on each of them. Such markers were used to define the reference coordinate system. In order to validate its robustness, statistical distribution of the point registration was then studied. Afterwards, three different observers marked the anatomical landmarks on a sub-sample of six specimens and the intra-observer and inter-observer variability of the point registration was performed.

RESULTS

Results of this study showed the highest values to be 1.46 mm (intra) and 4.08 mm (inter), both observed for the patellar ridge top. The intra-class correlation coefficient (ICC) for inter-observer variability ranked higher than 0.8 for all the landmarks used for the identification of the reference frame, and ranged from 0.4-0.9 for other landmarks.

CONCLUSIONS

This study demonstrates low intra-observer and inter-observer variability in the CT registration of landmarks that define and validate a robust coordinate system of the patella that could be used to perform accurate biomechanical and clinical studies.

The Effect of Patellar Tendon Release on the Characteristics of Patellofemoral Joint Squat Movement: A Simulation Analysis

Objectives: This paper studies the patellar tendon release’s effect on the movement characteristics of the artificial patellofemoral joint squat to provide reference data for knee joint surgery. Methods: Firstly, the dynamic finite element model of the human knee joint under squatting was established. Secondly, in the above no-release models, the release of 30% of the attachment area at the upper end, the lower end, or both ends of the patellar tendon were conducted, respectively. Then the simulations of all above four models were conducted. Finally, the results of the simulation were compared and analyzed. Results: The simulation results show that, after releasing the patellar tendon (compared with the no-release simulation’s results), the relative flexion, medial-lateral rotation, medial-lateral tilt, and superior-inferior shift of the patella relative to the femur increased; the medial-lateral shift and anterior-posterior shift of the patella relative to the femur decreased. Conclusion: In this paper, the maximum flexion angle of the patella increased after the patellar tendon being released (compared with the no-release model), which indicated that the mobility of knee joint was improved after the patellar tendon release. The simulation data in this paper can provide technical reference for total knee arthroplasty.

Tibio-femoral and patello-femoral joint kinematics during navigated total knee arthroplasty with patellar resurfacing

PURPOSE

In total knee arthroplasty, surgical navigation systems provide tibio-femoral joint (TFJ) tracking for relevant bone preparation, disregarding the patello-femoral joint (PFJ). Therefore, the important intra-operative assessment of the effect of component positioning, including the patella, on the kinematics of these two joints is not available. The objective of this study is to explore in vivo whether accurate tracking of the patella can result in a more physiological TFJ and PFJ kinematics during surgery.

METHODS

Ten patients underwent navigated knee replacement with patellar resurfacing. A secondary system was used to track patellar motion and PFJ kinematics using a special tracker. Patellar resection plane position and orientation were recorded using an instrumented probe. During all surgical steps, PFJ kinematics was measured in addition to TFJ kinematics.

RESULTS

Abnormal PFJ motion patterns were observed pre-operatively at the impaired knee. Patellar resection plane orientation on sagittal and transverse planes of 3.9° ± 9.0° and 0.4° ± 4.1° was found. A good restoration of both TFJ and PFJ kinematics was observed in all replaced knees after resurfacing, in particular the rotations in the three anatomical planes and medio-lateral patellar translation.

CONCLUSIONS

Patella tracking results in nearly physiological TFJ and PFJ kinematics in navigated knee arthroplasty with resurfacing. The intra-operative availability also of PFJ kinematics can support the positioning not only of the patellar component in case of resurfacing, but also of femoral and tibial components.

Physiological sagittal plane patellar kinematics during dynamic deep knee flexion

PURPOSE

Lateral radiographic views can be easily taken and have reveal considerable information about the patella. The purpose of this study was to obtain sagittal plane patellar kinematics data through the entire range of knee flexion under weight-bearing conditions.

METHODS

Patellar flexion angles relative to the femur and tibia and anterior-posterior and proximal-distal translations of the patella relative to the femur and tibia were measured from 0 to 165° knee flexion in nine healthy knees using dynamic radiographic images.

RESULTS

The patella flexed relative to the femur and tibia by two thirds times and one third times the knee flexion angle, respectively. The patella translated in an arc relative to the femur and tibia as the knee flexed. In early flexion, the superior and centroid points translated anteriorly and then the patella translated posteriorly relative to the femur. All three points of the patella translated posteriorly relative to the tibia during a full range of flexion. An average of four and three millimetres proximal patellar translation relative to the tibia was demonstrated from 0 to 20° and 140 to 160° knee flexion, respectively.

CONCLUSIONS

Physiological sagittal plane patellar kinematics, including patellar flexion angles and translations relative to the femur and tibia, showed generally similar patterns for each subject. Measurements of dynamic radiographic images under weight-bearing activities may enhance the opportunity to identify patellar pathological conditions.

A detailed and validated three dimensional dynamic model of the patellofemoral joint

A detailed 3D anatomical model of the patellofemoral joint was developed to study the tracking, force, contact and stability characteristics of the joint. The quadriceps was considered to include six components represented by 15 force vectors. The patellar tendon was modeled using four bundles of viscoelastic tensile elements. Each of the lateral and medial retinaculum was modeled by a three-bundle nonlinear spring. The femur and patella were considered as rigid bodies with their articular cartilage layers represented by an isotropic viscoelastic material. The geometrical and tracking data needed for model simulation, as well as validation of its results, were obtained from an in vivo experiment, involving MR imaging of a normal knee while performing isometric leg press against a constant 140 N force. The model was formulated within the framework of a rigid body spring model and solved using forth-order Runge-Kutta, for knee flexion angles between zero and 50 degrees. Results indicated a good agreement between the model predictions for patellar tracking and the experimental results with RMS deviations of about 2 mm for translations (less than 0.7 mm for patellar mediolateral shift), and 4 degrees for rotations (less than 3 degrees for patellar tilt). The contact pattern predicted by the model was also consistent with the results of the experiment and the literature. The joint contact force increased linearly with progressive knee flexion from 80 N to 210 N. The medial retinaculum experienced a peak force of 18 N at full extension that decreased with knee flexion and disappeared entirely at 20 degrees flexion. Analysis of the patellar time response to the quadriceps contraction suggested that the muscle activation most affected the patellar shift and tilt. These results are consistent with the recent observations in the literature concerning the significance of retinaculum and quadriceps in the patellar stability.

MRI of weight bearing and movement

Conventional, static magnetic resonance imaging (MRI) is able to provide a vast amount of information regarding the anatomy and pathology of the musculoskeletal system. However, patients, especially those whose pain is position dependent or elucidated by movement, may benefit from more advanced imaging techniques that allow for the acquisition of functional information. This manuscript reviews a variety of advancements in MRI techniques that are used to image the musculoskeletal system dynamically, while in motion or under load. The methodologies, advantages and drawbacks of stress MRI, cine-phase contrast MRI and real-time MRI are discussed as each has helped to advance the field by providing a scientific basis for understanding normal and pathological musculoskeletal anatomy and function. Advancements in dynamic MR imaging will certainly lead to improvements in the understanding, prevention, diagnosis and treatment of musculoskeletal disorders. It is difficult to anticipate that dynamic MRI will replace conventional MRI, however, dynamic MRI may provide additional valuable information to findings of conventional MRI.

Intraoperative 3D reconstruction of prostate brachytherapy implants with automatic pose correction

The success of prostate brachytherapy critically depends on delivering adequate dose to the prostate gland, and the capability of intraoperatively localizing implanted seeds provides potential for dose evaluation and optimization during therapy. REDMAPS is a recently reported algorithm that carries out seed localization by detecting, matching and reconstructing seeds in only a few seconds from three acquired x-ray images (Lee et al 2011 IEEE Trans. Med. Imaging 29 38-51). In this paper, we present an automatic pose correction (APC) process that is combined with REDMAPS to allow for both more accurate seed reconstruction and the use of images with relatively large pose errors. APC uses a set of reconstructed seeds as a fiducial and corrects the image pose by minimizing the overall projection error. The seed matching and APC are iteratively computed until a stopping condition is met. Simulations and clinical studies show that APC significantly improves the reconstructions with an overall average matching rate of ⩾99.4%, reconstruction error of ⩽0.5 mm, and the matching solution optimality of ⩾99.8%.

Three-dimensional patellar motion at the natural knee during passive flexion/extension. An in vitro study

Patellar maltracking may result in many patellofemoral joint (PFJ) disorders in the natural and replaced knee. The literature providing quantitative reference for normal PFJ kinematics according to which patellar maltracking could be identified is still limited. The aim of this study was to measure in vitro accurately all six-degrees-of-freedom of patellar motion with respect to the femur and tibia on 20 normal specimens. A state-of-the-art knee navigation system, suitably adapted for this study aim, was used. Anatomical reference frames were defined for the femur, tibia, and patella according to international recommendations. PFJ flexion, tilt, rotation, and translations were calculated in addition to standard tibiofemoral joint (TFJ) kinematics. All motion patterns were found to be generally repeatable intra-/interspecimens. PFJ flexion was 62% of the corresponding TFJ flexion range; tilt and translations along femoral mediolateral and tibial proximodistal axes during TFJ flexion were found with medial, lateral, and distal trends and within 12 degrees , 6 and 9 mm, respectively. No clear pattern for PFJ rotation was observed. These results concur with comparable reports from the literature and contribute to the controversial knowledge on normal PFJ kinematics. Their consistence provides fundamental information to understand orthopedic treatment of the knee and for possible relevant measurements intraoperatively.

Feasibility of using real-time MRI to measure joint kinematics in 1.5T and open-bore 0.5T systems

PURPOSE

To test the feasibility and accuracy of measuring joint motion with real-time MRI in a 1.5T scanner and in a 0.5T open-bore scanner and to assess the dependence of measurement accuracy on movement speed.

MATERIALS AND METHODS

We developed an MRI-compatible motion phantom to evaluate the accuracy of tracking bone positions with real-time MRI for varying movement speeds. The measurement error was determined by comparing phantom positions estimated from real-time MRI to those measured using optical motion capture techniques. To assess the feasibility of measuring in vivo joint motion, we calculated 2D knee joint kinematics during knee extension in six subjects and compared them to previously reported measurements.

RESULTS

Measurement accuracy decreased as the phantom's movement speed increased. The measurement accuracy was within 2 mm for velocities up to 217 mm/s in the 1.5T scanner and 38 mm/s in the 0.5T scanner. We measured knee joint kinematics with small intraobserver variation (variance of 0.8 degrees for rotation and 3.6% of patellar width for translation).

CONCLUSION

Our results suggest that real-time MRI can be used to measure joint kinematics when 2 mm accuracy is sufficient. They can also be used to prescribe the speed of joint motion necessary to achieve certain measurement accuracy.

Sensitivity of knee replacement contact calculations to kinematic measurement errors

The ability to measure in vivo knee kinematics accurately makes it tempting to calculate in vivo contact forces, pressures, and areas directly from kinematic data. However, the sensitivity of contact calculations to kinematic measurement errors has not been adequately investigated. To address this issue, we developed a series of sensitivity analyses derived from a validated in vivo computational simulation of gait. The simulation used an elastic foundation contact model to reproduce in vivo contact force, center of pressure, and fluoroscopic motion data collected from an instrumented knee replacement. Treating each degree of freedom (DOF) in the simulation as motion controlled, we first quantified how errors in measured relative pose of the implant components affected contact calculations. Pose variations of +/-0.1 mm or degree over the entire gait cycle changed maximum contact force, pressure, and area by 204, 100, and 117%, respectively. Larger variations of +/-0.5 mm or degree changed these same quantities by 1157, 108, and 578%, respectively. In both cases, the largest sensitivities were to errors in superior-inferior translation and varus-valgus rotation, with loss of contact occurring on one or both sides. We then quantified how switching the sensitive DOFs from motion to load control affected the sensitivity results. Pose variations of +/-0.5 mm or degree in the remaining DOFs changed maximum contact quantities by at most 3%. These results suggest that accuracy on the order of microns and milliradians is needed to estimate contact forces, pressures, and areas directly from in vivo kinematic measurements, and that use of load rather than motion control for the sensitive DOFs may improve the accuracy of in vivo contact calculations.

Patellar tracking during total knee arthroplasty: an in vitro feasibility study

Abnormal patellar tracking results in patello-femoral (PF) joint disorders and frequently in failure of total knee arthroplasty (TKA). It is fundamental to assess this tracking intra-operatively, i.e. since the implantation of the femoral and tibial components. The aim of this study was to assess the feasibility of three-dimensional anatomical-based patellar tracking intra-operatively in standard TKA. A surgical navigation system was utilized to test the new technique in-vitro. An original tracking device and a reference frame were designed and an articular convention for the description of PF joint kinematics was adopted. Six fresh-frozen amputated legs were analyzed with the new technique. Landmark digitations were used to define anatomical reference frames for the femur, tibia, and patella. Five trials of passive flexion were performed with 100 N force on the quadriceps, before and after standard knee arthroplasty. Patellar flexion, tilt, rotation and shift were calculated in addition to standard tibio-femoral (TF) joint kinematics. An intra-specimen repeatable path of motion over repetitions and a coupled path of motion throughout the flexion-extension cycle were observed in all intact knees, both at the TF and PF joints. Replication of the original PF motion in the intact knee was not fully accomplished in the replaced knee. These results revealed the feasibility and the necessity of patellar tracking during TKA. By monitoring intra-operatively also the PF kinematics, the surgeon has a more complete prediction of the performance of the final implant and therefore a valuable support for the most critical surgical decisions.

From scans to sutures: computer-assisted orthopedic surgery in the twenty-first century

Computer-assisted surgery is the process of using medical images, such as CT scans, X-ray fluoroscopy, or 3D ultrasound, to improve patient care. A typical surgical procedure begins by acquiring and processing a CT scan with specially developed image-analysis software. A surgeon then performs a "virtual surgery" on the patient to develop a preoperative plan. In the operating room the medical image is registered to the patient's anatomy by finding an optimal rigid-body transformation. This transformation allows an object or motion in one coordinate frame to be represented in the other frame, and thus a surgeon can visualize the location of an instrument deep within concealed anatomy while avoiding structures at risk. The operating surgeon can also use computer-tracked fluoroscopy or ultrasound for 3D guidance. For the past seven years, our interdisciplinary research group has been investigating fundamental problems in orthopedic surgery of bones and joints. This paper is an overview of the problems and solutions that have been tested in a set of pilot clinical trials in which we have treated more than 250 patients for early or advanced arthritis, poorly healed bone fractures, and treatment of deep bone tumors.

Validation of a low-dose hybrid RSA and fluoroscopy technique: Determination of accuracy, bias and precision

Analyzing skeletal kinematics with radiostereometric analysis (RSA) following corrective orthopedic surgery allows the quantitative comparison of different implant designs. The purpose of this study was to validate a technique for dynamically estimating the relative position and orientation of skeletal segments using RSA and single plane X-ray fluoroscopy. Two micrometer-based in vitro phantom models of the skeletal segments in the hip and knee joints were used. The spatial positions of tantalum markers that were implanted into each skeletal segment were reconstructed using RSA. The position and orientation of each segment were determined in fluoroscopy images by minimizing the difference between the markers measured and projected in the image plane. Accuracy was determined in terms of bias and precision by analyzing the deviation between the applied displacement protocol and measured pose estimates. Measured translational accuracy was less than 100 microm parallel to the image plane and less than 700 microm in the direction orthogonal to the image plane. The measured rotational error was less than 1 degrees . Measured translational and rotational bias was not statistically significant at the 95% level of confidence. The technique allows real-time kinematic skeletal measurements to be performed on human subjects implanted with tantalum markers for quantitatively measuring the motion of normal joints and different implant designs.

FTRAC--a robust fluoroscope tracking fiducial

C-arm fluoroscopy is ubiquitous in contemporary surgery, but it lacks the ability to accurately reconstruct three-dimensional (3D) information. A major obstacle in fluoroscopic reconstruction is discerning the pose of the x-ray image, in 3D space. Optical/magnetic trackers tend to be prohibitively expensive, intrusive and cumbersome in many applications. We present single-image-based fluoroscope tracking (FTRAC) with the use of an external radiographic fiducial consisting of a mathematically optimized set of ellipses, lines, and points. This is an improvement over contemporary fiducials, which use only points. The fiducial encodes six degrees of freedom in a single image by creating a unique view from any direction. A nonlinear optimizer can rapidly compute the pose of the fiducial using this image. The current embodiment has salient attributes: small dimensions (3 x 3 x 5 cm); need not be close to the anatomy of interest; and accurately segmentable. We tested the fiducial and the pose recovery method on synthetic data and also experimentally on a precisely machined mechanical phantom. Pose recovery in phantom experiments had an accuracy of 0.56 mm in translation and 0.33 degrees in orientation. Object reconstruction had a mean error of 0.53 mm with 0.16 mm STD. The method offers accuracies similar to commercial tracking systems, and appears to be sufficiently robust for intraoperative quantitative C-arm fluoroscopy. Simulation experiments indicate that the size can be further reduced to 1 x 1 X 2 cm, with only a marginal drop in accuracy.