Search

Previous studies have demonstrated a frequent occurrence of screw/K-wire malpositioning during surgical fracture treatment under 2D fluoroscopy and a correspondingly high revision rate as a result of using intraoperative 3D imaging. In order to facilitate and accelerate the diagnosis of implant malpositioning in 3D data sets, this study investigates two versions of an implant detection software for mobile 3D C-arms in terms of their detection performance based on comparison with manual evaluation. The 3D data sets of patients who had received surgical fracture treatment at five anatomical regions were extracted from the research database. First, manual evaluation of the data sets was performed, and the number of implanted implants was assessed. For 25 data sets, the time required by four investigators to adjust each implant was monitored. Subsequently, the evaluation was performed using both software versions based on the following detection parameters: true-positive-rate, false-negative-rate, false-detection-rate and positive predictive value. Furthermore, the causes of false positive and false negative detected implants depending on the anatomical region were investigated. Two hundred fourteen data sets with overall 1767 implants were included. The detection parameters were significantly improved (p<.001) from version 1 to version 2 of the implant detection software. Automatic evaluation required an average of 4.1±0.4 s while manual evaluation was completed in 136.15±72.9 s (p<.001), with a statistically significant difference between experienced and inexperienced users (p=.005). In summary, version 2 of the implant detection software achieved significantly better results. The time saved by using the software could contribute to optimizing the intraoperative workflow., (© 2022. The Author(s).)

Purpose: Reduction and osteosynthesis of ankle fractures is a challenging surgical procedure when it comes to the verification of the reduction result. Evaluation is conducted using intra-operative imaging of the injured ankle and depends on the expertise of the surgeon. Studies suggest that intra-individual variance of the ankle bone shape and pose is considerably lower than the inter-individual variance. It stands to reason that the information gain from the healthy contralateral side can help to improve the evaluation., Method: In this paper, an assistance system is proposed that provides a side-to-side view of the two ankle joints for visual comparison and instant evaluation using only one 3D C-arm image. Two convolutional neural networks (CNN) are employed to extract the relevant image regions and pose information of each ankle so that they can be aligned with each other. A first U-Net uses a sliding window to predict the location of each ankle. The standard plane estimation is formulated as segmentation problem so that a second U-Net predicts the three viewing planes for alignment., Results: Experiments were conducted to assess the accuracy of the individual steps on 218 unilateral ankle datasets as well as the overall performance on 7 bilateral ankle datasets. The experiments on unilateral ankles yield a median position-to-plane error of [Formula: see text] mm and a median angular error between 2.98[Formula: see text] and 3.71[Formula: see text] for the plane normals., Conclusion: Standard plane estimation via segmentation outperforms direct pose regression. Furthermore, the complete pipeline was evaluated including ankle detection and subsequent plane estimation on bilateral datasets. The proposed pipeline enables a direct contralateral side comparison without additional radiation. This has the potential to ease and improve the intra-operative evaluation for the surgeons in the future and reduce the need for revision surgery.

The aim of our study was to evaluate whether software-based artifact reduction can achieve an improved image quality, using intraoperative 3D imaging in spinal surgery. A total of 49 intraoperative 3D image datasets of patients, who underwent surgery with pedicle screw placement, were retrospectively evaluated. The visibility of anatomical structures and the diameter of the pedicle screws were examined, with and without the application of the artifact reduction software. All software prototypes can improve the visibility of anatomical structures (P < 0.01), except MAR (metal artifact reduction) combined with IRIS (iterative reconstruction in image space) (P = 0.04). The algorithms MAR and MAR-2 can reduce the blooming artifacts significantly (P < 0.01), but SL (Shepp & Logan) cannot (P = 0.08-0.988). In summary, software-based artifact reduction for intraoperative 3D datasets can improve the current image quality. Additional information regarding the implant placement and the fracture reduction is therefore generated for the surgeon.

Purpose: Guidance and quality control in orthopedic surgery increasingly rely on intra-operative fluoroscopy using a mobile C-arm. The accurate acquisition of standardized and anatomy-specific projections is essential in this process. The corresponding iterative positioning of the C-arm is error prone and involves repeated manual acquisitions or even continuous fluoroscopy. To reduce time and radiation exposure for patients and clinical staff and to avoid errors in fracture reduction or implant placement, we aim at guiding-and in the long-run automating-this procedure., Methods: In contrast to the state of the art, we tackle this inherently ill-posed problem without requiring patient-individual prior information like preoperative computed tomography (CT) scans, without the need of registration and without requiring additional technical equipment besides the projection images themselves. We propose learning the necessary anatomical hints for efficient C-arm positioning from in silico simulations, leveraging masses of 3D CTs. Specifically, we propose a convolutional neural network regression model that predicts 5 degrees of freedom pose updates directly from a first X-ray image. The method is generalizable to different anatomical regions and standard projections., Results: Quantitative and qualitative validation was performed for two clinical applications involving two highly dissimilar anatomies, namely the lumbar spine and the proximal femur. Starting from one initial projection, the mean absolute pose error to the desired standard pose is iteratively reduced across different anatomy-specific standard projections. Acquisitions of both hip joints on 4 cadavers allowed for an evaluation on clinical data, demonstrating that the approach generalizes without retraining., Conclusion: Overall, the results suggest the feasibility of an efficient deep learning-based automated positioning procedure, which is trained on simulations. Our proposed 2-stage approach for C-arm positioning significantly improves accuracy on synthetic images. In addition, we demonstrated that learning based on simulations translates to acceptable performance on real X-rays.

Purpose: Fracture reduction and fixation of syndesmotic injuries is a common procedure in trauma surgery. An intra-operative evaluation of the surgical outcome is challenging due to high inter-individual anatomical variation. A comparison to the contralateral uninjured ankle would be highly beneficial but would also incur additional radiation and time consumption. In this work, we pioneer automatic contralateral side comparison while avoiding an additional 3D scan., Methods: We reconstruct an accurate 3D surface of the uninjured ankle joint from three low-dose 2D fluoroscopic projections. Through CNN complemented 3D shape model segmentation, we create a reference model of the injured ankle while addressing the issues of metal artifacts and initialization. Following 2D-3D multiple bone reconstruction, a final reference contour can be created and matched to the uninjured ankle for contralateral side comparison without any user interaction., Results: The accuracy and robustness of individual workflow steps were assessed using 81 C-arm datasets, with 2D and 3D images available for injured and uninjured ankles. Furthermore, the entire workflow was tested on eleven clinical cases. These experiments showed an overall average Hausdorff distance of [Formula: see text] mm measured at clinical evaluation level., Conclusion: Reference contours of the contralateral side reconstructed from three projection images can assist surgeons in optimizing reduction results, reducing the duration of radiation exposure and potentially improving postoperative outcomes in the long term.

Medical images are essential in modern traumatology and orthopedic surgery. Access to images is often cumbersome due to a limited number of workstations. Moreover, due to the tremendous increase of data, the time to review or to communicate images has also become limited. One approach to overcome these problems is to make use of modern mobile devices, like tablet computers, to facilitate image access and associated workflows. Ten orthopedic surgeons were equipped with an Apple iPad mini 2 and specialized viewing software for medical images. The surgeons were able to send images from a workstation onto the tablets or to search for patient images directly. The software enabled the physicians to share images, annotated key slices, and messages instantly with their colleagues. The surgeons carried the tablets within or in the periphery of the hospital. The participants evaluated the software by means of daily questionnaires. Data was collected for a period of 9 months. Nearly 25 images were viewed in total by the surgeons per day. The tablet viewer was used for accessing approximately 30% of these images. On average, the surgeons were asked 1.7 times per day by a colleague for a second opinion. They used the tablets in approximately 29% of these cases. Furthermore, the mean time for accessing images was significantly lower using mobile software compared to conventional methods. Tablet computers can play a vital role for image access and communication in the daily routine of an orthopedic surgeon. Mobile image access is an important aspect for surgeons, especially in larger facilities, to facilitate and accelerate the clinical workflows.

Purpose: During a standard fracture reduction and fixation procedure of the distal radius, only fluoroscopic images are available for planning of the screw placement and monitoring of the drill bit trajectory. Our prototype intra-operative framework integrates planning and drill guidance for a simplified and improved planning transfer., Methods: Guidance information is extracted using a video camera mounted onto a surgical drill. Real-time feedback of the drill bit position is provided using an augmented view of the planning X-rays. We evaluate the accuracy of the placed screws on plastic bones and on healthy and fractured forearm specimens. We also investigate the difference in accuracy between guided screw placement versus freehand. Moreover, the accuracy of the real-time position feedback of the drill bit is evaluated., Results: A total of 166 screws were placed. On 37 plastic bones, our obtained accuracy was [Formula: see text] mm, [Formula: see text] and [Formula: see text] in tip position and orientation (azimuth and elevation), respectively. On the three healthy forearm specimens, our obtained accuracy was [Formula: see text] mm, [Formula: see text] and [Formula: see text]. On the two fractured specimens, we attained: [Formula: see text] mm, [Formula: see text] and [Formula: see text]. When screw plans were applied freehand (without our guidance system), the achieved accuracy was [Formula: see text] mm, [Formula: see text], while when they were transferred under guidance, we obtained [Formula: see text] mm, [Formula: see text]., Conclusions: Our results show that our framework is expected to increase the accuracy in screw positioning and to improve robustness w.r.t. freehand placement.

Purpose: In orthopedic surgeries, it is important to avoid intra-articular implant placements, which increase revision rates and the risk of arthritis. In order to support the intraoperative assessment and correction of surgical implants, we present an automatic detection approach using cone-beam computed tomography (CBCT)., Methods: Multiple active shape models (ASM) with specified articular surface regions are used to isolate the joint spaces. Fast and easy-to-implement methods are integrated in the ASM segmentation to optimize the robustness and accuracy for intraoperative application. A cylinder detection method is applied to determine metal implants. Intersections between articular surfaces and cylinders are detected and used to find intra-articular collisions., Results: Segmentations of two calcaneal articular surfaces were evaluated on 50 patient images and have shown average surface distance errors of 0.59 and 0.46 mm, respectively. The proposed model-independent segmentation at the specified articular surface regions allowed to significantly decrease the error by 22 and 25 % on average. The method was able to compensate suboptimal initializations for translations of up to 16 mm and rotations of up to 21[Formula: see text]. In a human cadaver test, articular perforations could be localized with an accuracy of 0.80 mm on average., Conclusions: A concept for automatic intraoperative detection of intra-articular implants in CBCT images was presented. The results show a reliable segmentation of articular surfaces in retrospective patient data and an accurate localization of misplaced implants in artificially created human cadaver test cases.

Purpose: The assessment of intra-operatively acquired volumetric data is a difficult and often time-consuming task, which demands a new set of skills from the surgeons. In the case of orthopedic surgeries such as the treatment of calcaneal fractures, the correctness of the reduction of the bone fragments can be verified with the help of C-arm CT volumetric images. For an accurate intra-operative assessment of the displaced fragments, an automatic segmentation of the articular surfaces and color-coded visualization was developed., Methods: Our automatic approach consists of three major steps: first, using adjusted standard planes intersecting the articular region, the joint space is localized with an intensity profile-based method. In a second step, the localized joint space is segmented on the Laplacian of Gaussian filtered volumetric image by a modified binary flood fill algorithm. Finally, a 3D surface model of the segmented joint space is analyzed and visualized with focus on critical displacements of the surface., Results: A specifically designed human cadaver study consisting of ten lower legs of ten different donors was conducted to acquire 48 realistic C-arm CT images of misaligned bone fragments (steps of varying sizes) in the posterior talar articular surface of the calcaneus. The proposed algorithmic pipeline was verified by the acquired image data and showed very good results with no false positives and an overall correct displacement assessment of 93.8%., Conclusions: The proposed algorithmic pipeline can be easily integrated into the clinical workflow and qualifies for intra-operative usage. It showed very good results on the reference data set of the cadaver study. With the help of such an assistance system, the time-consuming process of 2D view adjustment and visual assessment of the gray value images can be greatly simplified.

Purpose: With the help of an intra-operative mobile C-arm CT, medical interventions can be verified and corrected, avoiding the need for a post-operative CT and a second intervention. An exact adjustment of standard plane positions is necessary for the best possible assessment of the anatomical regions of interest but the mobility of the C-arm causes the need for a time-consuming manual adjustment. In this article, we present an automatic plane adjustment at the example of calcaneal fractures., Methods: We developed two feature detection methods (2D and pseudo-3D) based on SURF key points and also transferred the SURF approach to 3D. Combined with an atlas-based registration, our algorithm adjusts the standard planes of the calcaneal C-arm images automatically. The robustness of the algorithms is evaluated using a clinical data set. Additionally, we tested the algorithm's performance for two registration approaches, two resolutions of C-arm images and two methods for metal artifact reduction., Results: For the feature extraction, the novel 3D-SURF approach performs best. As expected, a higher resolution ([Formula: see text] voxel) leads also to more robust feature points and is therefore slightly better than the [Formula: see text] voxel images (standard setting of device). Our comparison of two different artifact reduction methods and the complete removal of metal in the images shows that our approach is highly robust against artifacts and the number and position of metal implants., Conclusions: By introducing our fast algorithmic processing pipeline, we developed the first steps for a fully automatic assistance system for the assessment of C-arm CT images.

Purpose: Orthopedic fractures are often fixed using metal implants. The correct positioning of cylindrical implants such as surgical screws, rods and guide wires is highly important. Intraoperative 3D imaging is often used to ensure proper implant placement. However, 3D image interaction is time-consuming and requires experience. We developed an automatic method that simplifies and accelerates location assessment of cylindrical implants in 3D images., Methods: Our approach is composed of three major steps. At first, cylindrical characteristics are detected by analyzing image gradients in small image regions. Next, these characteristics are grouped in a cluster analysis. The clusters represent cylindrical implants and are used to initialize a cylinder-to-image registration. Finally, the two end points are optimized regarding image contrast along the cylinder axis., Results: A total of 67 images containing 420 cylindrical implants were used for testing. Different anatomical regions (calcaneus, spine) and various image sources (two mobile devices, three reconstruction methods) were investigated. Depending on the evaluation set, the detection performance was between 91.7 and 96.1% true- positive rate with a false-positive rate between 2.0 and 3.2%. The end point distance errors ranged from [Formula: see text] to [Formula: see text] mm and the orientation errors from [Formula: see text] to [Formula: see text] degrees. The average computation time was less than 5 seconds., Conclusions: An automatic method was developed and tested that obviates the need for 3D image interaction during intraoperative assessment of cylindrical orthopedic implants. The required time for working with the viewing software of cone-beam CT device is drastically reduced and leads to a shorter time under anesthesia for the patient.

Purpose: Segmentation of the proximal femur in digital antero-posterior (AP) pelvic radiographs is required to create a three-dimensional model of the hip joint for use in planning and treatment. However, manually extracting the femoral contour is tedious and prone to subjective bias, while automatic segmentation must accommodate poor image quality, anatomical structure overlap, and femur deformity. A new method was developed for femur segmentation in AP pelvic radiographs., Methods: Using manual annotations on 100 AP pelvic radiographs, a statistical shape model (SSM) and a statistical appearance model (SAM) of the femur contour were constructed. The SSM and SAM were used to segment new AP pelvic radiographs with a three-stage approach. At initialization, the mean SSM model is coarsely registered to the femur in the AP radiograph through a scaled rigid registration. Mahalanobis distance defined on the SAM is employed as the search criteria for each annotated suggested landmark location. Dynamic programming was used to eliminate ambiguities. After all landmarks are assigned, a regularized non-rigid registration method deforms the current mean shape of SSM to produce a new segmentation of proximal femur. The second and third stages are iteratively executed to convergence., Results: A set of 100 clinical AP pelvic radiographs (not used for training) were evaluated. The mean segmentation error was 0.96 mm ± 0.35 mm, requiring <5 s per case when implemented with Matlab. The influence of the initialization on segmentation results was tested by six clinicians, demonstrating no significance difference., Conclusions: A fast, robust and accurate method for femur segmentation in digital AP pelvic radiographs was developed by combining SSM and SAM with dynamic programming. This method can be extended to segmentation of other bony structures such as the pelvis.

We propose a new method for fully-automatic landmark detection and shape segmentation in X-ray images. Our algorithm works by estimating the displacements from image patches to the (unknown) landmark positions and then integrating them via voting. The fundamental contribution is that, we jointly estimate the displacements from all patches to multiple landmarks together, by considering not only the training data but also geometric constraints on the test image. The various constraints constitute a convex objective function that can be solved efficiently. Validated on three challenging datasets, our method achieves high accuracy in landmark detection, and, combined with statistical shape model, gives a better performance in shape segmentation compared to the state-of-the-art methods.