Design of an ultrasound-emitting drill guide for freehand pedicle screw navigation.

BACKGROUND: Accurate pedicle screw placement in spinal surgery is critical as inaccuracies can lead to morbidity and suboptimal outcomes. Navigation and robotics have reduced malplacement rates, but their adoption is limited by high costs, learning curves, surgical time, and radiation. The authors propose an ultrasound-emitting and self-localising drill guide for precise screw placement that overcomes the limitations of current techniques. MATERIALS AND METHODS: The preliminary configuration analysis involves systematically varying design parameters and assessing localization performance using lumbar spine MRI based simulations. The authors evaluate localization techniques based on accuracy and optimization capture range. RESULTS: Results suggest that feasible designs can accurately estimate position. A promising design features a 5 mm radius cannula with ten 35mm-long ultrasound strips, 32 elements per strip, and a fanned-out emission profile. A multi-start active-set optimization algorithm with six initial estimates ensures reliable and efficient localization. CONCLUSIONS: The simulation suggests that the proposed design can achieve sufficient localization accuracy for pedicle screw navigation. These findings will guide the fabrication of a novel ultrasound-emitting drill guide for further evaluation and physical testing.

Using acoustic feedback and analyses for removal of bone cement during revision knee replacement surgeries.

PURPOSE: The purpose of this study was to analyze the scraping sounds generated during revision knee replacement surgeries to discriminate between the inner cortical bone and the cement, with the goal of minimizing bone removal and increasing the structural integrity of the revision. METHODS: We prepared seven porcine femurs by partially filling them with bone cement, and recorded scraping sounds produced by a surgical scraping tool. We used a hierarchical machine learning approach to first detect a contact and then classify it as either bone or cement. This approach was based on a Support Vector Machine learning algorithm that was fed with temporal and spectral features of the sounds. A Leave-One-Bone-Out validation method was used to assess the performance of the proposed method. RESULTS: The average recall for the noncontact, bone, and cement classes was 98%, 75%, and 72%, respectively. The corresponding precision for the respective classes was 99%, 67%, and 61%. CONCLUSION: The scraping sound that is generated during revision replacement surgeries carries significant information about the material that is being scraped. Such information can be extracted using a supervised machine learning algorithm. The scraping sound produced during revision replacement procedures can potentially be used to enhance cement removal during knee revision surgery. Future work will assess whether such monitoring can increase the structural integrity of the revision.

What are the optimal targeting visualizations for performing surgical navigation of iliosacral screws? A user study.

INTRODUCTION: Complex orthopaedic procedures, such as iliosacral screw (ISS) fixations, can take advantage of surgical navigation technology to achieve accurate results. Although the impact of surgical navigation on outcomes has been studied, no studies to date have quantified how the design of the targeting display used for navigation affects ISS targeting performance. However, it is known in other contexts that how task information is displayed can have significant effects on both accuracy and time required to perform motor tasks, and that this can be different among users with different experience levels. This study aimed to investigate which visualization techniques helped experienced surgeons and inexperienced users most efficiently and accurately align a surgical tool to a target axis. METHODS: We recruited 21 participants and conducted a user study to investigate five proposed 2D visualizations (bullseye, rotated bullseye, target-fixed, tool-fixed in translation, and tool-fixed in translation and rotation) with varying representations of the ISS targets and tool, and one 3D visualization. We measured the targeting accuracy achieved by each participant, as well as the time required to perform the task using each of the visualizations. RESULTS: We found that all 2D visualizations had equivalent translational and rotational errors, with mean translational errors below 0.9 mm and rotational errors below 1.1[Formula: see text]. The 3D visualization had statistically greater mean translational and rotational errors (4.29 mm and 5.47[Formula: see text], p < 0.001) across all users. We also found that the 2D bullseye view allowed users to complete the simulated task most efficiently (mean 30.2 s; 95% CI 26.4-35.7 s), even when combined with other visualizations. CONCLUSIONS: Our results show that 2D bullseye views helped both experienced orthopaedic trauma surgeons and inexperienced users target iliosacral screws accurately and efficiently. These findings could inform the design of visualizations for use in a surgical navigation system for screw insertions for both training and surgical practice.

Can Technology Assistance be Cost Effective in TKA? A Simulation-Based Analysis of a Risk-prioritized, Practice-specific Framework.

BACKGROUND: Robotic, navigated, and patient-specific instrumentation (PSI) TKA procedures have been introduced to improve component placement precision and improve implant survivorship and other clinical outcomes. However, the best available evidence has shown that these technologies are ineffective in reducing revision rates in the general TKA patient population. Nonetheless, it seems plausible that these technologies could be an effective and cost-effective means of reducing revision risk in clinical populations that are at an elevated risk of revision because of patient-specific demographics (such as older age at index surgery, elevated BMI, and being a man). Since clinical trials on this topic would need to be very large, a simulation approach could provide insight on which clinical populations would be the most promising for analysis. QUESTIONS/PURPOSES: We conducted a simulation-based analysis and asked: (1) Given key demographic parameters characterizing a patient population, together with estimates of the precision achievable with selected forms of technology assistance in TKA, can we estimate the expected distributions of anticipated reductions in lifetime revision risk for that population and the associated improvements in quality-adjusted life years (QALYs) that would be expected to result? (2) Are there realistic practice characteristics (such as combinations of local patient demographics and capital and per-procedure costs) for which applying a per-patient risk-prioritized policy for using technology-assisted TKA could be considered cost-effective based on projected cost savings from reductions in revision rates? METHODS: We designed simulations of hypothetical practice-specific clinical scenarios, each characterized by patient volume, patient demographics, and technology-assisted surgical technique, using demographic information drawn from other studies to characterize two contrasting simulated clinical scenarios in which the distributions of factors describing patients undergoing TKA place one population at a comparatively elevated risk of revision (elevated-risk population) and the second at a comparatively reduced risk of revision (lower-risk population). We used results from previous systematic reviews and meta-analyses to estimate the implant precision in coronal plane alignment for patient-specific instrumentation, navigated, and robotic technology. We generated simulated TKA patient populations based on risk estimates from large clinical studies, structured reviews, and meta-analyses and calculated the patient-specific reduction in the revision risk and the change in QALYs attributable to the technology-assisted intervention in each of the two simulated clinical scenarios. We also incorporated a sensitivity analysis, incorporating variations in the effect size of deviations from overall coronal alignment on revision risk and difference in health state utilities acquired through a structured review process. We then simulated the outcomes of 25,000 operations per patient using the precisions associated with the conventional TKA technique, the three technology-assisted techniques, and a hypothetical technology-assisted intervention that could consistently deliver perfectly neutral overall coronal alignment, which is unachievable in practice. A risk-prioritized treatment policy was emulated by ordering the simulated patients from the highest to lowest predicted increase in QALYs, such that simulated patients who would see the greatest increase in the QALYs (and therefore the greatest reduction in lifetime revision risk) were the patients to receive technology-assisted TKA intervention in a practice. We used cost estimates acquired through a structured review process and calculated the net added costs of each of the three technology-assisted techniques as a function of the percent utilization (proportion of patients treated with technology assistance in a practice), factoring in fixed costs, per-procedure variable costs, and savings occurring from the prevention of future revision surgery. Finally, we calculated the incremental cost-effectiveness ratio (ICER) and marginal cost-effectiveness ratio (MCER) for each technology-assisted technique for the two clinical scenarios. We then used a Monte Carlo approach to simulate variations in key patient risk, health state, and economic factors as well as to obtain a distribution of estimates for cost-effectiveness. We considered an intervention to be cost effective if either the ICER or MCER values were below USD/QALY 63,000. RESULTS: For the lower-risk population, the median reduction in the revision risk was 0.9% (0.4% to 2.2%, extrema from the sensitivity analysis) and 1.8% (0.9% to 4.4%) for PSI and robotic TKA, respectively, and 1.9% (1.0% to 4.6%) for ideal TKA. In contrast, the median reduction in the revision risk in the elevated-risk clinical scenario was 2.0% (1.2% to 3.4%) and 4.6% (2.7% to 8.5%) for PSI and robotic TKA and 5.1% (3.0% to 9.4%) for ideal TKA. Estimated differences in the cumulative gain in QALYs attributable to technology-assisted TKA ranged from 0.6 (0.2 to 1.8) to 4.0 (1.8 to 10.0) QALYs per 100 patients, depending on the intervention type and clinical scenario. For PSI, we found treating 15% of patients in the lower-risk population and 77% in the elevated-risk population could meet the threshold for being considered cost effective. For navigated TKA systems offering high alignment precision, we found the intervention could meet this threshold for practice sizes of at least 300 patients per year and a percent utilization of 27% in the lower-risk population. In the elevated-risk population, cost-effectiveness could be achieved in practice volumes as small as 100 patients per year with a percent utilization of at least 6%, and cost savings could be achieved with a percent utilization of at least 45%. We found that robotic TKA could only meet the threshold for being considered cost-effectiveness in the lower-risk population if yearly patient volumes exceeded 600 and for a limited range of percent utilization (27% to 32%). However, in the elevated-risk patient population, robotic TKA with high alignment precision could potentially be cost effective for practice sizes as small as 100 patients per year and a percent utilization of at least 20% if a risk-prioritized treatment protocol were used. CONCLUSION: Based on these simulations, a selective-use policy for technology-assisted TKA that prioritizes using technology assistance for those patients at a higher risk of revision based on patient-specific factors could potentially meet the cost-effectiveness threshold in selected circumstances (for example, primarily in elevated-risk populations and larger practice sizes). Whether it does meet that threshold would depend significantly on the surgical precision that can be achieved in practice for a given proposed technology as well as on the true local costs of using the proposed technology. We further recommend that any future randomized trials seeking to demonstrate possible effects of technology assistance on revision risk focus on clinical populations that are at higher risk of revision (such as, patient populations that are relatively younger, have higher BMIs, and higher proportions of men). CLINICAL RELEVANCE: This study suggests that technology assistance is only likely to prove cost effective in selected circumstances rather than in all clinical populations and practice settings. In general, we project that surgical navigation is most likely to prove cost effective in the widest range of circumstances, that PSI may be cost effective or cost neutral in a moderate range of circumstances, and that robotic surgery is only likely to be cost effective in moderately large practices containing patients who are on average at an intrinsically elevated risk of revision.

Can uncertainty estimation predict segmentation performance in ultrasound bone imaging?

PURPOSE: Segmenting bone surfaces in ultrasound (US) is a fundamental step in US-based computer-assisted orthopaedic surgeries. Neural network-based segmentation techniques are a natural choice for this, given promising results in related tasks. However, to gain widespread use, we must be able to know how much to trust segmentation networks during clinical deployment when ground-truth data is unavailable. METHODS: We investigated alternative ways to measure the uncertainty of trained networks by implementing a baseline U-Net trained on a large dataset, together with three uncertainty estimation modifications: Monte Carlo dropout, test time augmentation, and ensemble learning. We measured the segmentation performance, calibration quality, and the ability to predict segmentation performance on test data. We further investigated the effect of data quality on these measures. RESULTS: Overall, we found that ensemble learning with binary cross-entropy (BCE) loss achieved the best segmentation performance (mean Dice: 0.75-0.78 and RMS distance: 0.62-0.86mm) and the lowest calibration errors (mean: 0.22-0.28%). In contrast to previous studies of area or volumetric segmentation, we found that the resulting uncertainty measures are not reliable proxies for surface segmentation performance. CONCLUSION: Our experiments indicate that a significant performance and confidence calibration boost can be achieved with ensemble learning and BCE loss, as tested on 13,687 US images containing various anatomies and imaging parameters. However, these techniques do not allow us to reliably predict future segmentation performance. The results of this study can be used to improve the calibration and performance of US segmentation networks.

Targeting repeatability of a less obtrusive surgical navigation procedure for total shoulder arthroplasty.

PURPOSE: Surgical navigation systems have demonstrated improvements in alignment accuracy in a number of arthroplasty procedures, but they have not yet been widely adopted for use in total shoulder arthroplasty (TSA). We believe this is due in part to the obtrusiveness of conventional optical tracking systems, as well as the need for additional intraoperative steps such as calibration and registration. The purpose of this study is to evaluate the feasibility of adapting a less-intrusive dental navigation system for use in TSA. METHODS: We developed a proof-of-concept system based on validated laser-engraved surgical tools recently introduced for use in dental surgery that are calibrated once when manufactured and not recalibrated at time of use. The design also features a notably smaller bone-mounted tracker that can be tracked from a wide range of viewing angles. To assess our system's performance, we modified the dental surgical software to support guidance of a TSA procedure. We then conducted a user study in which three participants with varying surgical experience used the system to drill 30 holes in a glenoid model. Using a coordinate measuring machine, we determined the resulting drilled trajectory and compared this to the pre-planned trajectory. Since we used a model glenoid rather than anatomical specimens, we report on targeting precision rather than overall procedure precision or accuracy. RESULTS: We found targeting precision < 1 mm (standard deviation) for locating the entry hole and < ~ 1° (SD) for both version and inclination. The latter value was markedly lower than the end-to-end angular precision achieved by previously reported TSA navigation systems (approximately 3°-5° SD). CONCLUSION: We conclude that variability during the targeting phase represents a small fraction of the overall variability exhibited by existing systems, so a less obtrusive navigation system for TSA based on laser-engraved tooling is likely feasible, which could improve the uptake rates of surgical navigation for TSA, thereby potentially leading to improved overall surgical outcomes.

How Large a Study Is Needed to Detect TKA Revision Rate Reductions Attributable to Robotic or Navigated Technologies? A Simulation-based Power Analysis.

BACKGROUND: Robotic and navigated TKA procedures have been introduced to improve component placement precision in the hope of improving implant survivorship and other clinical outcomes. Although numerous comparative studies have shown enhanced precision and accuracy in placing components, most comparative studies have not shown that such interventions result in improved implant survival. Given what we know about effect sizes from large arthroplasty registries, large cohort studies, and large randomized controlled trials (RCTs), we wondered how large randomized trials would need to be to detect such small differences, and if the number is very high, what that would tell us about the value of these treatments for preventing revision surgery. QUESTIONS/PURPOSES: In this simulation study, we asked: Given that survivorship differences between technology-assisted TKA (TA-TKA, which we defined as either navigated or robot-assisted TKA) and conventional TKA are either small or absent based on large arthroplasty registries, large cohort studies, and large RCTs, how large would randomized trials need to be to detect small differences between TA-TKA and conventional TKA if they exist, and how long would the follow-up period need to be to have a reasonable chance to detect those differences? METHODS: We used estimated effect sizes drawn from previous clinical and registry studies, combined with estimates of the accuracy and precision of various navigation and robotic systems, to model and simulate the likely outcomes of potential comparative clinical study designs. To characterize the ranges of patients enrolled and general follow-up times associated with traditional RCT studies, we conducted a structured search of previously published studies evaluating the effect of robotics and navigation on revision rates compared with that of conventional TKA. The structured search of the University of British Columbia's library database (which automatically searches medical publication databases such as PubMed, Embase, Medline, and Web of Science) and subsequent searching through included studies' reference lists yielded 103 search results. Only clinical studies assessing implant survival differences between patient cohorts of TA-TKA and conventional TKA were included. Studies analyzing registry data, using cadaver specimens, assessing revision TKA, conference proceedings, and preprint services were excluded. Twenty studies met all our inclusion criteria, but only one study reported a statistically significant difference between the conventional and robotic or navigated groups. Next, we generated a large set of patients with simulated TKA (1.5 million), randomly assigning each simulated patient a set of patient-specific factors (age at the index surgery, gender, and BMI) drawn from data from registries and published information. We divided this set of simulated procedures into four groups, each associated with a coronal alignment precision reported for different types of surgical procedures, and randomly assigned each patient an overall coronal alignment consistent with their group's precision. TA procedures were modeled based on the alignment precision that an intervention could deliver, regardless of whether the technology used was navigation- or robot-assisted. To evaluate the power associated with using different cohort sizes, we ran a Monte Carlo simulation generating 3000 simulated populations that were drawn (with replacement) from the large set of simulated patients with TKA. We simulated the time to revision for aseptic loosening for each patient, computed the corresponding Kaplan-Meier survival curves, and applied a log-rank test to each study for statistical differences in revision rates at concurrent follow-up timepoints (1-25 years). From each simulation associated with a given cohort size, we determined the percentage of simulated studies that found a statistically significant difference at each follow-up interval. For each alternative precision, we then also calculated the expected reduction in revision rates (effect size) attributable to TA-TKA intervention and the number needed to treat (NNT) using TA-TKA to prevent one revision at 2, 5, 10, and 15 years after index surgery for the entire set of Kaplan-Meier survival analyses. RESULTS: The results from our simulation found survivorship differences favoring TA-TKA ranging from 1.4% to 2.0% at 15 years of follow-up. Comparative studies would need to enroll between 2500 and 4000 patients in each arm of the study, depending on the precision of the navigated or robotic procedure, to have an 80% chance of showing this reduction in revision rates at 15 years of follow-up. For the highest precision simulated intervention, the NNT using TA-TKA to prevent one revision was 1000 at 2 years, 334 at 5 years, 100 at 10 years, and 50 at 15 years post-index surgery. CONCLUSION: Based on these simulations, it appears that TA-TKA interventions could potentially result in a relative reduction in revision rates as large as 27% (from 7.5% down to about 5.5% at 15 years for the intervention with the most precise coronal alignment); however, since this 2% absolute reduction in revision rates is relatively small in comparison with the baseline success rate of TKA and would not be realized until 15 years after the index surgery, traditional RCT studies would require excessively large numbers of patients to be enrolled and excessively long follow-up times to demonstrate whether such a reduction actually exists. CLINICAL RELEVANCE: Given that the NNTs to avoid revisions at various time points are predicted to be high, it would require correspondingly low system costs to justify broad adoption of TA-TKA based on avoided revision costs alone, though we speculate that technology assistance could perhaps prove to be cost effective in the care of patients who are at an elevated risk of revision.

Automatically Delineating Key Anatomy in 3-D Ultrasound Volumes for Hip Dysplasia Screening.

Developmental dysplasia of the hip (DDH) metrics based on 3-D ultrasound have proven more reliable than those based on 2-D images, but to date have been based mainly on hand-engineered features. Here, we test the performance of 3-D convolutional neural networks for automatically segmenting and delineating the key anatomical structures used to define DDH metrics: the pelvis bone surface and the femoral head. Our models are trained and tested on a data set of 136 volumes from 34 participants. For the pelvis, a 3D-U-Net achieves a Dice score of 85%, outperforming the confidence-weighted structured phase symmetry algorithm (Dice score = 19%). For the femoral head, the 3D-U-Net had centre and radius errors of 1.42 and 0.46 mm, respectively, outperforming the random forest classifier (3.90 and 2.01 mm). The improved segmentation may improve DDH measurement accuracy and reliability, which could reduce misdiagnosis.

Deep learning-based X-ray inpainting for improving spinal 2D-3D registration.

BACKGROUND: Two-dimensional (2D)-3D registration is challenging in the presence of implant projections on intraoperative images, which can limit the registration capture range. Here, we investigate the use of deep-learning-based inpainting for removing implant projections from the X-rays to improve the registration performance. METHODS: We trained deep-learning-based inpainting models that can fill in the implant projections on X-rays. Clinical datasets were collected to evaluate the inpainting based on six image similarity measures. The effect of X-ray inpainting on capture range of 2D-3D registration was also evaluated. RESULTS: The X-ray inpainting significantly improved the similarity between the inpainted images and the ground truth. When applying inpainting before the 2D-3D registration process, we demonstrated significant recovery of the capture range by up to 85%. CONCLUSION: Applying deep-learning-based inpainting on X-ray images masked by implants can markedly improve the capture range of the associated 2D-3D registration task.

3-D Ultrasound Imaging Reliability of Measuring Dysplasia Metrics in Infants.

Developmental dysplasia of the hip is a hip abnormality that ranges from mild acetabular dysplasia to irreducible femoral head dislocations. While 2-D B-mode ultrasound (US)-based dysplasia metrics or disease metrics are currently used clinically to diagnose developmental dysplasia of the hip, such estimates suffer from high inter-exam variability. In this work, we propose and evaluate 3-D US-derived dysplasia metrics that are automatically computed and demonstrate that these automatically derived dysplasia metrics are considerably more reproducible. The key features of our automatic method are (i) a random forest-based learning technique to remove regions across the coronal axis that do not contain bone structures necessary for dysplasia-metric extraction, thereby reducing outliers; (ii) a bone segmentation method that uses rotation-invariant and intensity-invariant filters, thus remaining robust to signal dropout and varying bone morphology; (iii) a novel slice-based learning and 3-D reconstruction strategy to estimate a probability map of the hypoechoic femoral head in the US volume; and (iv) formulae for calculating the 3-D US-derived dysplasia metrics. We validate our proposed method on real clinical data acquired from 40 infant hip examinations. Results show a considerable (around 70%) reduction in variability in two key 3-D US-derived dysplasia metrics compared with their 2-D counterparts.

Ultrasound Bone Segmentation: A Scoping Review of Techniques and Validation Practices.

Ultrasound bone segmentation is an important yet challenging task for many clinical applications. Several works have emerged attempting to improve and automate bone segmentation, which has led to a variety of computational techniques, validation practices and applied clinical scenarios. We characterize this exciting and growing body of research by reviewing published ultrasound bone segmentation techniques. We review 56 articles in detail and categorize and discuss the image analysis techniques that have been used for bone segmentation. We highlight the general trends of this field in terms of clinical motivation, image analysis techniques, ultrasound modalities and the types of validation practices used to quantify segmentation performance. Finally, we present an outlook on promising areas of research based on the unaddressed needs for solving ultrasound bone segmentation.

Using Laser Range-finding to Measure Bore Depth in Surgical Drilling of Bone.

BACKGROUND: Measuring drilled bore depth in bone is an important part of osteosynthesis surgery. Current methods have substantial limitations in terms of reliability, leading to placement of incorrectly sized screws and unsatisfactory user experience. QUESTIONS/PURPOSES: (1) Can a prototype laser range-finder measure bore depth in bone as well as or better than a conventional depth gauge in terms of accuracy and precision (that is, variability)? METHODS: A conventional analog orthopaedic surgical depth gauge was compared with a laser range-finder-based prototype. Experiments were conducted on four pig hind limbs, with bicortical holes drilled in the femur and the tibia. Two surgeons alternated drilling bores in three different clinically relevant conditions: straight drilling through the diaphysis, angled drilling through the diaphysis, and straight drilling through the metaphysis. Depth measurements were taken with the laser range-finder-based prototype, and the conventional depth gauge and compared against depth measurements obtained from a CT image that served as the reference measurement. RESULTS: In straight diaphyseal drilling the laser range-finder-based prototype had a larger mean error of 1.34 mm (± 0.7 mm) compared with a mean error of -0.06 mm (± 1.38 mm) using the conventional gauge (95% CI 0.824 to 1.976; p < 0.001). In angled diaphyseal drilling, there was no difference in mean error between the laser range-finder-based prototype (1.66 ± 0.86 mm) and the conventional gauge (2.36 ± 3.79 mm [95% CI -2.338 to 0.938]; p = 0.393). In straight metaphyseal drilling, there was no difference in mean error between the laser range-finder-based prototype (2.11 ± 0.8 mm) and the conventional gauge (1.51 ± 3.19 mm [95% CI -0.500 to 1.700]; p = 0.280). The laser range-finder-based prototype had greater precision (smaller variance) than the conventional depth gauge in straight diaphyseal drilling (p < 0.001), angled diaphyseal drilling (p < 0.001), and straight metaphyseal drilling (p < 0.001). CONCLUSIONS: A laser range-finder-based prototype mounted on a conventional surgical drill demonstrated overall similar accuracy and better precision in measuring drilled bore depth in bone compared with the conventional depth gauge. CLINICAL RELEVANCE: A device based on this concept could improve the reliability of bore depth measurement in surgical practice and could therefore reduce the frequency of screw replacement and associated complications.

A comparative analysis of intensity-based 2D-3D registration for intraoperative use in pedicle screw insertion surgeries.

PURPOSE: Although multiple algorithms have been reported that focus on improving the accuracy of 2D-3D registration techniques, there has been relatively little attention paid to quantifying their capture range. In this paper, we analyze the capture range for a number of variant formulations of the 2D-3D registration problem in the context of pedicle screw insertion surgery. METHODS: We tested twelve 2D-3D registration techniques for capture range under different clinically realistic conditions. A registration was considered as successful if its error was less than 2 mm and 2° in 95% of the cases. We assessed the sensitivity of capture range to a variety of clinically realistic parameters including: X-ray contrast, number and configuration of X-rays, presence or absence of implants in the image, inter-subject variability, intervertebral motion and single-level vs multi-level registration. RESULTS: Gradient correlation + Powell optimizer had the widest capture range and the least sensitivity to X-ray contrast. The combination of 4 AP + lateral X-rays had the widest capture range (725 mm2). The presence of implant projections significantly reduced the registration capture range (up to 84%). Different spine shapes resulted in minor variations in registration capture range (SD 78 mm2). Intervertebral angulations of less than 1.5° had modest effects on the capture range. CONCLUSIONS: This paper assessed capture range of a number of variants of intensity-based 2D-3D registration algorithms in clinically realistic situations (for the use in pedicle screw insertion surgery). We conclude that a registration approach based on the gradient correlation similarity and the Powell's optimization algorithm, using a minimum of two C-arm images, is likely sufficiently robust for the proposed application.

Fast and automatic bone segmentation and registration of 3D ultrasound to CT for the full pelvic anatomy: a comparative study.

PURPOSE: Ultrasound (US) is a safer alternative to X-rays for bone imaging, and its popularity for orthopedic surgical navigation is growing. Routine use of intraoperative US for navigation requires fast, accurate and automatic alignment of tracked US to preoperative computed tomography (CT) patient models. Our group previously investigated image segmentation and registration to align untracked US to CT of only the partial pelvic anatomy. In this paper, we extend this to study the performance of these previously published techniques over the full pelvis in a tracked framework, to characterize their suitability in more realistic scenarios, along with an additional simplified segmentation method and similarity metric for registration. METHOD: We evaluated phase symmetry segmentation, and Gaussian mixture model (GMM) and coherent point drift (CPD) registration methods on a pelvic phantom augmented with human soft tissue images. Additionally, we proposed and evaluated a simplified 3D bone segmentation algorithm we call Shadow-Peak (SP), which uses acoustic shadowing and peak intensities to detect bone surfaces. We paired this with a registration pipeline that optimizes the normalized cross-correlation (NCC) between distance maps of the segmented US-CT images. RESULTS: SP segmentation combined with the proposed NCC registration successfully aligned tracked US volumes to the preoperative CT model in all trials, in contrast to the other techniques. SP with NCC achieved a median target registration error (TRE) of 2.44 mm (maximum 4.06 mm), when imaging all three anterior pelvic structures, and a mean runtime of 27.3 s. SP segmentation with CPD registration was the next most accurate combination: median TRE of 3.19 mm (maximum 6.07 mm), though a much faster runtime of 4.2 s. CONCLUSION: We demonstrate an accurate, automatic image processing pipeline for intraoperative alignment of US-CT over the full pelvis and compare its performance with the state-of-the-art methods. The proposed methods are amenable to clinical implementation due to their high accuracy on realistic data and acceptably low runtimes.

A deep learning framework for segmentation and pose estimation of pedicle screw implants based on C-arm fluoroscopy.

PURPOSE: Pedicle screw fixation is a challenging procedure with a concerning rates of reoperation. After insertion of the screws is completed, the most common intraoperative verification approach is to acquire anterior-posterior and lateral radiographic images, based on which the surgeons try to visually assess the correctness of insertion. Given the limited accuracy of the existing verification techniques, we identified the need for an accurate and automated pedicle screw assessment system that can verify the screw insertion intraoperatively. For doing so, this paper offers a framework for automatic segmentation and pose estimation of pedicle screws based on deep learning principles. METHODS: Segmentation of pedicle screw X-ray projections was performed by a convolutional neural network. The network could isolate the input X-rays into three classes: screw head, screw shaft and background. Once all the screw shafts were segmented, knowledge about the spatial configuration of the acquired biplanar X-rays was used to identify the correspondence between the projections. Pose estimation was then performed to estimate the 6 degree-of-freedom pose of each screw. The performance of the proposed pose estimation method was tested on a porcine specimen. RESULTS: The developed machine learning framework was capable of segmenting the screw shafts with 93% and 83% accuracy when tested on synthetic X-rays and on clinically realistic X-rays, respectively. The pose estimation accuracy of this method was shown to be [Formula: see text] and [Formula: see text] on clinically realistic X-rays. CONCLUSIONS: The proposed system offers an accurate and fully automatic pedicle screw segmentation and pose assessment framework. Such a system can help to provide an intraoperative pedicle screw insertion assessment protocol with minimal interference with the existing surgical routines.

A Systematic Review and Meta-analysis on the Reproducibility of Ultrasound-based Metrics for Assessing Developmental Dysplasia of the Hip.

BACKGROUND: The purposes of this study were to (1) perform a systematic review of articles that reported agreement or reproducibility in repeated diagnosis of developmental dysplasia of the hip (DDH) using ultrasound imaging, (2) estimate the reproducibility in the available dysplasia metrics, and (3) compare reproducibility of the available dysplasia metrics. METHODS: A systematic review of the Medline and Embase databases was performed by using a search strategy formulated from our research question: "For infants at risk of DDH, are US imaging-based diagnoses reproducible?" Two reviewers independently identified articles for inclusion in the systematic review, and then assessed the quality of the included studies using the Guidelines for Reporting Reliability and Agreement Studies guideline. Variability and agreement-related statistics in the included studies were extracted and included in a meta-analysis for summarizing the available statistics. The reproducibility of the available dysplasia metrics was compared, with a Bonferroni correction made to adjust for multiple comparisons. RESULTS: Twenty eight studies were included in the systematic review. Overall, the quality of the included studies was moderate (average, 10.7/15; range, 6 to 12). Graf's alpha angle had the lowest interexamination variability of the metrics assessed, followed by Graf's beta angle (the variability of the alpha angle was 10% lower than the variability of the beta angle, P<0.05). However, despite Graf's angles having lower variability compared with other dysplasia metrics, their actual variability was still problematically high. This finding was supported by the low intraclass correlation and Kappa coefficient values reported in the included studies. There was also evidence to suggest that the reproducibility in DDH diagnosis has potentially worsened over time. CONCLUSIONS: Overall, we found high variability and low agreement in all reported dysplasia metrics. Furthermore, in the last 3 decades, the repeatability of dysplasia metrics has not markedly improved and may even have declined, indicating a genuine need for improving repeatability and reliability of ultrasound-based DDH diagnosis. LEVEL OF EVIDENCE: Level III-systematic review of level III studies.

An intraoperative fluoroscopic method to accurately measure the post-implantation position of pedicle screws.

PURPOSE: Pedicle screw malplacement, leading to neurological symptoms, vascular injury, and premature implant loosening, is not uncommon and difficult to reliably detect intraoperatively with current techniques. We propose a new intraoperative post-placement pedicle screw position assessment system that can therefore allow surgeons to correct breaches during the procedure. Our objectives were to assess the accuracy and robustness of this proposed screw location system and to compare its performance to that of 2D planar radiography. METHODS: The proposed system uses two intraoperative X-ray shots acquired with a standard fluoroscopic C-arm and processed using 2D/3D registration methods to provide a 3D visualization of the vertebra and screw superimposed on one another. Point digitization and CT imaging of the residual screw tunnel were used to assess accuracy in five synthetic lumbar vertebral models (10 screws in total). Additionally, the accuracy was evaluated with and without correcting for image distortion and for various screw lengths, screw materials, breach directions, and vertebral levels. RESULTS: The proposed method is capable of localizing the implanted screws with less than 2 mm of translational error (RMSE: 0.7 and 0.8 mm for the screw head and tip, respectively) and less than [Formula: see text] angular error (RMSE: [Formula: see text]), with minimal change to the errors if image distortion is not corrected. Breaches and their anatomical locations were all correctly visualized and identified for a variety of screw lengths, screw materials, breach locations, and vertebral levels, demonstrating the robustness of this approach. In contrast, one breach, one non-breach, and the anatomical location of three screws were misclassified with 2D X-ray. CONCLUSION: We have demonstrated an accurate and low-radiation technique for localizing pedicle screws post-implantation that requires only two X-rays. This intraoperative feedback of screw location and direction may allow the surgeon to correct malplaced screws intraoperatively, thereby reducing postoperative complications and reoperation rates.

Automatic Evaluation of Scan Adequacy and Dysplasia Metrics in 2-D Ultrasound Images of the Neonatal Hip.

Ultrasound (US) imaging of an infant's hip joint is widely used for early detection of developmental dysplasia of the hip. In current US-based diagnosis of developmental dysplasia of the hip, trained clinicians acquire US images and, if they judge them to be adequate (i.e., to contain relevant hip joint structures), analyze them manually to extract clinically useful dysplasia metrics. However, both the scan adequacy classification and dysplasia metrics extraction steps exhibit significant variability within and between both clinicians and institutions, which can result in significant over- and undertreatment rates. To reduce the subjectivity resulting from this variability, we propose a computational image analysis technique that automatically identifies adequate images and subsequently extracts dysplasia metrics from these 2-D US images. Our automatic method uses local phase symmetry-based image measures to robustly identify intensity-invariant geometric features of bone/cartilage boundaries from the US images. Using the extracted geometric features, we trained a random forest classifier to classify images as adequate or inadequate, and in the adequate images we used a subset of the geometric features to calculate key dysplasia metrics. We validated our method on a data set of 693 US scans collected from 35 infants. Our approach produces excellent agreement with clinician adequacy classifications (area under the receiver operating characteristic curve = 0.985) and in reducing variability in the measured developmental dysplasia of the hip metrics (p < 0.05). The automatically computed dysplasia metrics appear to be slightly biased toward higher Graf categories than the manually estimated metrics, which could potentially reduce missed early diagnoses.

Automatic extraction of bone surfaces from 3D ultrasound images in orthopaedic trauma cases.

PURPOSE: 3D ultrasound (US) imaging has the potential to become a powerful alternative imaging modality in orthopaedic surgery as it is radiation-free and can produce 3D images (in contrast to fluoroscopy) in near-real time. Conventional B-mode US images, however, are characterized by high levels of noise and reverberation artifacts, image quality is user-dependent, and bone surfaces are blurred, which makes it difficult to both interpret images and to use them as a basis for navigated interventions. 3D US has great potential to assist orthopaedic care, possibly assisting during surgery if the anatomical structures of interest could be localized and visualized with sufficient accuracy and clarity and in a highly automated rapid manner. METHODS: In this paper, we present clinical results for a novel 3D US segmentation technique we have recently developed based on multi-resolution analysis to localize bone surfaces in 3D US volumes. Our method is validated on scans obtained from 29 trauma patients with distal radius and pelvic ring fractures. RESULTS: Qualitative and quantitative results demonstrate remarkably clear segmentations of bone surfaces with an average surface fitting error of 0.62 mm (standard deviation (SD) of 0.42 mm) for pelvic patients and 0.21 mm (SD 0.14 mm) for distal radius patients. CONCLUSIONS: These results suggest that our technique is sufficiently accurate for potential use in orthopaedic trauma applications.