CAOS 2024: Papers with Abstracts

Abstract. Traditionally, surgeons measure lower limb deformities manually by assessing angles between axes drawn on full lower limb X-rays connecting specific landmarks. This process is considered cumbersome and subject to the surgeon's expertise. Our study aims to alleviate the manual detection of landmarks while enhancing the assessment of lower limb malalignment through an innovative approach that combines coordinate regression and landmark segmentation. While various deep learning solutions exist, our method differs by using landmark segmentation to indicate the possible location of the landmarks; this information is combined with the X-rays to estimate the position of the landmarks via coordinate regression. We named this deep learning architecture segmentation-guided regression.
To address the performance of our proposed approach, we evaluated the detection errors for eight landmarks and measured five malalignment metrics. We also compare our approach against landmark regression and landmark segmentation. While landmark segmentation achieved accurate landmark identification, it faced challenges in malalignment measurement due to incorrectly detected landmarks. On the other hand, regression had no failed detections but exhibited lower landmark detection accuracy. Our segmentation-guided regression showed a balance, ensuring no incorrect landmark detections, improved landmark accuracy, and precise malalignment quantification.
By encouraging the coordinate regression network to focus on specific areas through segmentation guidance, our method positions landmarks more accurately and effectively measures malalignment. Consequently, our approach provides surgeons with a reliable tool for comprehensive lower limb malalignment assessment, combining the strengths of coordinate regression and landmark segmentation.

Abstract. This study investigated the impact of (1) the transition from an MR workflow to a GB workflow as well as (2) the adoption of the GB workflow on the bone cut parameters during TKA. The findings demonstrated that both the transition and the adoption have an impact on the bone cut parameters and surgeons seeking to switch from MR to GB surgical workflow should be aware of these changes.

Abstract. Recent alignment techniques for total knee arthroplasty (TKA) allow the possibility of targeting knee joint laxity goals at the planning stage. In the absence of clear guidance in terms of the definition of these goals, surgeons tend to set-up the laxity goals based on their individual preferences. This retrospective review based on the same knee system and the same surgical technique established that the laxity goals were found to be surgeon specific.

Abstract. About 4 % of THA patients suffer from iliopsoas syndrome (IS). Modelling of the iliopsoas muscles is considered to be a reasonable aspect for identification of risk factors for IS. In this study we modelled the m. iliacus and the m. psoas major as separate muscles, consisting of 3 fascicles each. We analysed length and angular change of the fascicles from preoperative to postoperative state in a reference position on a database of 201 Japanese patients that underwent unilateral primary THA, with five of these patients being diagnosed with IS postoperatively. We did not find any correlation between length and angular changes and the appearance of IS. Further investigation is needed to find relevant risk factors for prediction of IS.

Abstract. Definition of target zones for prosthesis alignment is a common method in preoperative planning for THA. Several criteria for calculation of these target zones are discussed in literature and a combined target zone (CTZ) has been defined on the basis of morphological and morphofunctional parameters. Especially the definition of a patient-specific target ROM is important for prediction of postoperative prosthesis ́ functionality. Prosthetic impingement can limit movements und thus restrict patients in their activities of daily living. A precise analysis of the postoperative target ROM can help to provide prosthetic impingement during movements, which are essential for the individual patient and also improve the size of the resulting CTZ for prosthesis alignment. We analyzed four different ROM definitions and their impact on the size of the ROM based target zone (ROMTZ) as part of the CTZ on a database of 200 patients. We found significant differences between the sizes of the resulting ROMTZs. Hence, our study underlines the effectiveness of a differentiated evaluation of the individual target ROM for optimized preoperative planning in THA.

Abstract. Deep learning techniques for diagnosing Developmental Dysplasia of the Hip (DDH) in newborns from ultrasound (US) images of the hip have demonstrated improved reliability over manual annotations of US scans. While volumetric 3D US has been shown to better represent hip bone morphology, most of the proposed automatic diagnostic approaches to measure 3D equivalents of the commonly used 2D Graf angle rely on strong morphological (geometric) priors. We have found that a significant fraction of cases (~20%) result in metrics which expert assessors regard as incorrect or implausible. We hypothesize that the lack of robustness of existing algorithms is due to their assumption that selected morphological priors are always valid, and this may not hold in a number of cases. In this study, we evaluate the differences between extracted DDH metrics based on expert labels and automatic segmentations. We show that a metric extraction process that uses morphological priors is sensitive to relatively small variations in the segmentation results.

Abstract. Measuring the knee laxities when performing total knee arthroplasty (TKA) is part of the regular workflow and can be used as input for intra-operative planning. Gaps are traditionally measured by varus-valgus tests, but new instrumented methods have recently been proposed. This cadaveric study showed that the instrumented method is more reproducible than the manual one.

Abstract. New technologies introduction in the operating room induces a cost for the health system which can be assessed. This evaluation should focus on the stages dedicated to this innovation, not on the whole surgical workflow. The aim of this study was to evaluate the learning time for surgeons using a new intraoperative planning technique coupled with instrumented knee laxity measurement.

Abstract. For computer-assisted percutaneous scaphoid fixation, a patient-specific bone model is required for surgical planning. This bone model is commonly derived from pre- operative computed tomography (CT) or magnetic resonance imaging (MRI) data. We propose an approach for bone model derivation based on intra-operative 3D ultrasound (US) imaging for the cases, where pre-operative diagnostic CT or MRI are not indicated or available. As scaphoid bone surfaces are only partially visible in sonographic images, we employ the Transformer-based AdaPoinTr architecture to incorporate statistical morphological knowledge for the completion of partial bone surfaces extracted from sonographic images. For the generation of datasets, we built a statistical shape model (SSM) based on 85 scaphoid bone models. From this SSM, we generated 12288 full scaphoid models for training. 20 additional scaphoid bone models were used for testing. Partial models for both training and testing were generated by subsampling the full models, mimicking 3D US imaging from a volar probe position. Evaluation of the final trained model on the test subset showed a mean symmetric distance of 0.3 mm between original and completed scaphoid models, with an inference time of 0.2 s per model. We furthermore planned screws based on the completed test models and evaluated their fit for the original models. We found no screw protrusion for any tested model, with a mean safety margin to bone surface of 0.7 mm. This study shows feasibility of our approach for US-based bone model generation; future work may aim at integrating dorsal surface information.

Abstract. Introduction: The adoption of new technology should be supported by improvements in patient-reported outcomes (PROMs). The purpose of this study was to assess the one-year PROMs of patients who underwent total hip arthroplasty (THA) using a novel, fluoroscopy-based, robotic-assisted (RA-THA) system when compared to a manual, fluoroscopic-assisted technique (mTHA).

Materials and methods: A review of 91 consecutive mTHA and 85 consecutive RA-THA via a direct anterior approach was conducted. All cases were performed by the same surgeon at the same institution, for a pre-operative diagnosis of osteoarthritis, avascular necrosis, or rheumatoid arthritis. Outcomes included one-year Veterans RAND-12 (VR-12) Physical/Mental, Hip Disability and Osteoarthritis Outcome (HOOS) Pain/Physical Function/Joint Replacement, and University of California Los Angeles (UCLA) Activity scores, as well as the difference between pre-operative and one-year post-operative PROMs.

Results: Patients in the RA-THA cohort had lower pre-operative HOOS-JR scores compared to patients in the mTHA cohort (37.0 vs. 43.1; p=0.031). Cohorts experienced similar one-year post-operative VR-12, HOOS, and UCLA Activity scores. Patients in the RA-THA cohort experienced greater improvements across all pre- and post-operative HOOS scores compared to patients in the mTHA cohort: Pain (+54.7 vs. +42.1; p=0.009), Physical Function (-41.6 vs. -28.7; p=0.007), and Joint Replacement (+46.6 vs. +33.0; p=0.002). These differences exceeded minimum clinically important difference (MCID).

Conclusions: Both manual and robotic cohorts experienced benefit from THA at one-year post-operative. Importantly, the use of a novel, fluoroscopy-based robotic assistance system for primary THA resulted in greater improvements in PROMs at one-year relative to manual technique.

Abstract. Introduction of deep-learning (DL) in the computer assisted surgery requires to train and evaluate segmentation models by maximizing the control and knowledge of data. In this study, we highlight the incomes of data mastery through the examples of shoulder bone segmentations.

Abstract. Complex bone deformation of the femur can occur in bone disease. This can cause significant functional impairment and is often an indication for surgery. Current technology enables preoperative planning and execution of these complex multi-planar osteotomies with high accuracy. A novel solution is presented that generates preoperative plans using Blender. The model automatically optimizes femoral shaft shape within the boundaries of clinical constraints. 20 cases of femoral deformity were processed by the model retrospectively. Eighteen of twenty preoperative plans were accepted by two independent clinical experts. This shows the ability of the model to generate preoperative plans of sufficient clinical quality. Four random cases were compared to manual preoperative planning. These showed comparable results. The collum anteversion angle was normalized in four cases in the automatic group and only in one for the manual group. The manual planning process was slower in every case, ranging from 31% to 575% slower. The proposed method, allows fast interactive surgical planning within a constraint based design and is a promising aide for complex femoral correction planning.

Abstract. Morpho-functional interbone parameters of the knee are often used in clinical practice to assess the functional anatomy of the individual patient. Respective parameters, such as the TT-TG distance, are regularly measured manually on CT or MRI images. To overcome this time-consuming process, an automated framework for knee morphological analysis was previously presented by our group. A relevant remaining limitation, however, is the imaging performed in supine and not in an active, weight-bearing position, which was addressed in this study.
Data from 7 patients (14 knees) scheduled for total hip arthroplasty were used for this study. After segmentation, the CT-derived bone surface models were matched to the EOS images based on landmarks, with manual control and optional correction. Subsequently, the automated framework for morphological analysis was applied.
A statistically significant difference in mean was found for the joint rotation. The other mean deviations were not statistically significant and below the parameters’ standard deviation, whereas the absolute deviations were higher. This fact highlights the relevance of inter-individual differences in supine versus standing interbone parameter measurements.
In general, large discrepancies regarding the changes in interbone parameter measurements from supine to standing position were found in the literature. The results of this study are plausible with regard to theoretical biomechanical relations. Overall, the study motivates an interbone parameter assessment in a standardized, active weight-bearing position. With further automation in data pre-processing, the workflow could be applied to large databases and hence be used to define reference ranges of interbone parameters in various poses.

Abstract. Purpose: This study aimed to investigate the effectiveness of tract-specific diffusion tensor imaging (DTI) metrics in identifying the responsible segments for neurological dysfunction in cervical spondylotic myelopathy (CSM).
Methods: The study encompassed nineteen patients in CSM group and ten healthy caregivers in control group. All participants underwent a comprehensive physical examination, MRI assessment, and DTI examination conducted by a senior chief physician. Intraoperative spinal cord electrophysiological examination was performed as the objective measure of spinal cord function during surgery for patients. MRI parameters including the aspect ratio, transverse ratio, T2 hyperintensities of the spinal cord and DTI metrics parameters such as axial diffusivity (AD), mean diffusivity (MD), radial diffusivity (RD), fractional anisotropy (FA) for both whole spinal cord column and dorsal column were collected. Receiver operating characteristic curves were constructed to evaluate the diagnostic efficacy of the parameters. The area under the curve (AUC), sensitivity, and specificity were calculated.
Results: The AUC of DTI (Dorsal column) parameters was larger than the corresponding parameters of the DTI (Whole spinal cord). AD of DTI-Dorsal Column possessed the greatest efficacy (AUC=0.823, sensitivity=84.21%, specificity=77.32%) to determine the responsible segment, larger than AD of DTI-Whole spinal cord, aspect ratio and transverse ratio. Diagnostic efficacy of DTI and MRI parameters was influenced by cervical spine segment.
Conclusions: AD from the DTI-Dorsal Column exhibited the most significant potential in identifying responsible segments. The diagnostic effectiveness of both DTI and MRI parameters was notably influenced by the specific cervical spine segment.

Abstract. Accurate glenoid component positioning in shoulder arthroplasty is important to avoid potential impingement, loosening, and instability. Several techniques are currently utilized to assist in glenoid guide pin positioning, although no studies exist that directly compare the accuracy between these techniques. The objective of this study was to compare guide pin insertion accuracy using traditional 3D software planning (TSP), patient specific instrumentation (PSI) guides, computer-navigation (C- NAV), and mixed reality navigation (MR-NAV).
Twenty shoulder computer tomography scans exhibiting glenohumeral arthritis or rotator cuff tear arthropathy were preoperatively planned for reverse shoulder arthroplasty. Quadruplicate models of each glenoid were plastic 3D printed and were used to randomly assess four guide pin insertion techniques by a fellowship trained surgeon as follows: (1) TSP, (2) PSI guides, (3) C-NAV, and (4) MR-NAV. Following guide pin placement, the absolute error in guide pin position and orientation relative to the preoperative plan was measured using a digitization system.
Similar inclination (P>0.066) and version (P>0.515) accuracy occurred between PSI, C-NAV, and MR-NAV techniques. Furthermore, all three methods exhibited significantly less error in guide pin inclination compared to TSP (P<0.025). Greater version error was also observed with TSP (4±3°) but was not significantly greater than the other techniques (P>0.063). The error in guide pin entry point was similar between all four methods utilized (P>0.086).
This study showed that the accuracy of PSI, C-NAV, and MR-NAV are superior to TSP for glenoid pin insertion in-vitro. Further investigation is needed to validate the accuracy of all guide pin insertion techniques in-vivo.

Abstract. Musculoskeletal disorders (MSDs) pose significant healthcare challenges. This study ad- dresses diagnostic limitations by proposing a novel algorithm able to estimate key anatom- ical features of the femur from thigh skin. Leveraging a dataset of 50 angioscanner thigh- femur pairs, we pioneer a robust Statistical Shape Model (SSM) that captures correlations between the skin and the underlying femur. Femur inference from a known thigh uses a Bayesian approach and demonstrates promising results. Although the femur reconstruc- tion error may seem high, it is important to note that the goal is MSK feature estimation rather than precise bone reconstruction. Preliminary results are promising, suggesting a potential application in non-invasive MSK diagnosis for surgery.

Abstract. Navigated orthopaedic surgery relies on bony landmarks and the accuracy of their acquisition can impact the surgery outcomes. We propose an automatic workflow to de- termine 11 femoral bony landmarks on virtual 3D meshes.
The studied landmarks were first determined on the mean shape of a statistical shape model of the femur. Then the statistical shape model was fitted to the virtual 3D meshes, and the landmarks of the mean shape were projected onto the fitted mesh. The proposed method was validated by comparing the computed landmarks to ground truth landmarks acquired manually on 41 knees. We also investigated the impact of the landmarks’ accu- racy on the variability of axes and resection planes derived from the considered landmarks. The 11 femoral bony landmarks were automatically determined in less than 2 minutes with an accuracy of 2.81 ± 1.86mm. Such error impacted the accuracy of the derived axes and planes with less than 0.5° angular deviation.
Three landmarks had poorer accuracy and precision attesting how ambiguous their defini- tion is and the difficulty to identify them. The proposed method allows the fast acquisition of femoral bony landmarks, with similar accuracy to manual approaches.

Abstract. Knee joint laxity or instability post total knee arthroplasty (TKA) is a common reason for patient dissatisfaction, and in some cases, can even result in the need for revision TKA. With the ultimate goal of developing an EOS-based knee joint laxity technique, the objective of this study was therefore to determine how repeatably EOS biplanar X-rays can measure 3D knee joint transforms on a radiographically realistic knee phantom between the femur and tibia in the presence of a TKA implant.
To assess repeatability, we first scanned a femoral and a tibial anatomical model (Tactile KneeTM, Tactile Orthopaedics) using a clinical CT scanner and segmented the scans semi-automatically. The fully assembled TKA phantom was then placed within a jig that maintained a fixed and rigid connection between the femur and tibia. This model was then scanned 10 times with the EOS system, with approximately equally spaced perturbations ranging up to ±300 around the superior/inferior axis, applied manually between scans. We then implemented a 2D-3D registration technique to assign Euler coordinate systems to both the femoral and tibial phantoms using JointTrack Auto. This process was completed for both the femoral and tibial TKA phantoms, and coordinate system assignment repeatability was calculated.
The mean deviations from the mean for the medial-lateral, anterior-posterior, and superior-inferior translational axes were all submillimetric: 0.6 ± 0.7 mm, 0.4 ± 0.6 mm, and 0.7 ± 0.8 mm, respectively. The mean deviation from the mean around the medial- lateral, anterior-posterior, and superior-inferior rotational axes were all on the order of 1 degree or less: 1.0 ± 1.20, 0.8 ± 1.10, and 1.0 ± 1.20, respectively.

Abstract. Given evidence that patients with a lower age at index surgery and higher BMI have an enhanced risk of aseptic revision. In this study we used a previously developed methodology to estimate how large a comparative study would need to be to detect differences between navigated/robotic TKA and conventional TKA in both higher- and lower-risk patient populations, and how long the follow-up periods would need to be to be sufficiently powered to detect those differences.
We modeled and simulated the likely outcomes of potential RCT study designs according to our previously published method. We generated three large sets of patients with distributions of patient-specific factors characteristic of patients at a reduced or enhanced risk of aseptic revision (relative to the typical risk assumed in our previous study). We then 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.
The results from our simulation found survivorship differences favoring TA-TKA ranging from 2.8% to 3.9% at 15- and 20-years follow-up on the patient population at an enhanced risk of aseptic revision. Even for the patient population at the highest baseline risk of aseptic revision, comparative studies would still need to enroll at least 1750 patients in each arm of the study to have an 80% chance of showing this reduction in revision rates at 15 years of follow-up. Traditional RCT studies would require impractically large numbers of patients to be enrolled and excessively long follow-up times to demonstrate whether such a reduction actually exists.

Abstract. Methods
We retrospectively collected 51 shoulder CT scans from 44 patients. Two senior and one junior orthopaedic shoulder surgeons created manual annotations of the glenoid best-fit circle and the maximum distance between the best-fit circle and the glenoid contour. Computed GBL annotations and measurements were obtained with our method. The GBL % ratio was computed from the resulting measurements.
The inter-observer variability on the surgical decision of Bankart vs. Latrajet with the 13.5% % ratio cut-off threshold was determined as follows. Each surgeon chose for each scan the required surgery based on their manual annotations. The agreement/disagreement between the surgeons and the effect of replacing the threshold with an interval was then derived. The effect of the selection of the best-fit circle and of the largest radial distance between the glenoid contour and the best-fit circle was determined by presenting four options to the surgeons: the manual annotations of each surgeon and the computed annotation with their respective GBL deficiency %. Each surgeon then chose one of the annotations.
Results
In the 20 cases in which the GBL % ratio was < 5% or > 25%, all three surgeons agreed on the surgical procedure. For the remaining 31 cases, they disagreeded in 13 cases (42%). When the GBL cut-off of 13.5% was replaced by the interval 12.0-16.5%, the disagreement disappeared. In only ~30% of the cases, the surgeons chose their own annotation. The computed annotations were selected by all at least as often as those of the junior surgeon.
Conclusion
The established GBL % ratio cut-off of 13.5% may lead to significant discrepancies between surgeons regarding the type of surgery to be performed. Replacing it by the interval 12.0-16.5% may improve decision making by helping identify borderline cases for which there is no consensus. The computed GBL % ratio is within the observer variability and may thus be reliably used to save time and increase decision consistency.

Abstract. Assessment of the loss in skeletal muscle mass (SMM) is often quantified by measuring the skeletal muscle index (SMI) through dual-energy x-ray absorptiometry (DXA). However, as SMI measurement is not always performed in the clinical setting, we aimed to develop and verify a system that predicts SMM loss from the CT images of the lower extremity that were acquired for the preoperative assessment of hip surgery.
Sixty female patients with hip diseases who underwent CT and a whole-body DXA scan were retrospectively analyzed. Using an automated CT segmentation model of each muscle, muscles of the lower extremities were segmented and classified into three groups: gluteal muscles, muscles above the knee joint, and all lower extremity muscles. The relationship between the total muscle mass of the three groups and DXA-measured lean mass was assessed. Further, the SMI of the three groups was calculated, and their diagnostic performance in predicting sarcopenia (SMI < 5.4 kg/m2) was evaluated using the receiver operating characteristic curve (ROC) analysis.
Strong correlations were observed with lean mass and SMM of gluteal muscles (r = 0.82, p < 0.01), above the knee joint (r = 0.90, p < 0.01), and of all lower extremities (r = 0.94, p < 0.01). The area under the curves for each region were 0.93, 0.90, and 0.96, respectively.
Our findings suggest that lower extremity CT scans with limited regions of interest may serve as a viable method for predicting whole-body SMM loss, indicating the possibility of diagnosing sarcopenia from such CT images.

Abstract. Surgical intervention is often indicated for complex multiplanar foot deformities (CMFDs). Addressing the 3D nature of these deformities proves challenging for achieving optimal foot alignment and reoperations are often necessary. Virtual surgical planning (VSP) can aid in the understanding of the multiplanar deformity and determine the osteotomies needed for correction. This study introduces a patient-specific guide (PSG) applied to a CMFD patient undergoing midfoot closing wedge osteotomy. The PSG design allowed for accurate reconstruction of the angular deformity, while allowing for intra-operative translational adjustments. Further improvement of the PSG design was proposed to ensure an even better fit in future applications.

Abstract. Predicting plantar pressure using personalized finite element (FE) biomechanical models of the patient’s anatomy holds significant promise for aiding in virtual surgical planning and predicting clinical outcomes. While sophisticated FE models have been described in literature, they are not suitable for routine clinical care as the personalization is complex and the model calculation is computationally intensive. Therefore, an automatically generated FE biomechanical model of the foot is presented, offering an efficient alternative for clinical application, and it is qualitatively compared against a computationally intensive sophisticated FE model previously published by our group.

Abstract. Three-dimensional printing technology has been developed and applied in orthopedic surgery, including knee arthroplasty. To match individual knee morphology, surgeons can choose between standardized implants or customized 3D-printed ones. Despite technical advancements, the routine adoption of 3D-printed implants faces slow progress and various barriers. Our study explores surgeons' perspectives on 3D-printed prostheses by inviting them to freely express their thoughts by answering the question “What do you think about the manufacture of a prosthesis by 3D printing?
The questionnaire was completed by 90 surgeons, with an average experience exceeding 10 years (57.8% ± 10.2%). The majority worked in public hospitals (60% ± 10.1%) and performed zero to 100 prostheses annually (66.7% ± 9.7%). A large proportion do not use planning software (52.2% ± 9.7%), navigation systems, or robots (68.9% ± 9.6%). Respondents collectively agreed on the additional surgical time necessitated by technological innovation (74.4% ± 9.0%). Opinions on 3D printing were diverse, with 70%± 9.5% expressing positivity and 30%± 9.5% holding negative views. Motivations were categorized into seven domains, centered around "pre-surgery" and "post-surgery" concerns. Notably, results showed that a favorable disposition towards 3D printing was associated with the use of navigation systems or robots.
In conclusion, while no outright opposition to implementation was found, some surgeons expressed a preference for validated results and raised concerns about the entire supply chain, encompassing hospitals, insurance companies, and manufacturers. Despite the absence of opposition, the full adoption of 3D printing in joint replacement hinges on advancements across various areas of joint surgery.

Abstract. To address the increasing global demand for Total Knee Arthroplasty and reduce the need for revisions, several technologies combining 3D planning and artificial intelligence have emerged. These innovations aim to enhance customization, improve component positioning accuracy and precision. The integration of these advancements paves the way for the development of personalized and connected knee implant.
These groundbreaking advancements may necessitate changes in surgical practices. Hence, it is important to comprehend surgeons' intentions in integrating these technologies into their routine procedures. Our study aims to assess how surgeons' preferences will affect the acceptability of using this new implant and associated technologies within the entire care chain.
We employed a Discrete Choice Experiment, a predictive technique mirroring real-world healthcare decisions, to assess surgeons' trade-off evaluations and preferences.
A total of 90 experienced surgeons, performing a significant number of procedures annually (mostly over 51) answered. Analysis indicates an affinity for technology but limited interest in integrating digital advancements like preoperative software and robotics. However, they are receptive to practice improvements and considering the adoption of future sensors.
In conclusion, surgeons prefer customized prostheses via augmented reality, accepting extra cost. Embedded sensor technology is deemed premature by them;

Abstract. This study explores the use of automated unsupervised evaluations via wearable devices, to assess the success of hip and knee replacements as a complement to traditional PROMs.
A comprehensive analysis was conducted on data from 1144 TKA and THA patients utilizing a mobile application, with activity data collected through the Garmin Vivofit 4 wearable device. Key parameters, including daily Peak 6-Minute Consecutive Cadence (P6MC) and daily Peak 1-Minute Cadence (P1M), were computed pre and post surgery and analyzed to assess the efficacy of these metrics in monitoring the recovery progress and the surgery outcomes.
Cadence measurements, specifically P6MC and P1M, emerged as robust indicators. These metrics exhibited a superior level of responsiveness compared to traditional step- count measurements and showed good complementarity with PRO’s traditionally used in clinical practices. Moreover, the capture of these parameters being daily, unsupervised, and automated gives the potential of offering more granularity and better compliance than PROMs, providing new insights to assess quality of new surgical techniques. Moreover, the growing ubiquity of smartphones and wearables makes the use of such metrics usable in daily practice.

Abstract. New surgical techniques, such as robotics and augmented reality, promise to increase the quality and reduce the variability of the surgical procedure. But the effect of these new developments could be altered by the variability of the post operative care. Rehabilitation following total knee replacement (TKA) traditionally involves in-person therapy, presenting challenges in terms of protocol adherence, reproducibility, and cost. Digital rehabilitation holds promise in addressing these issues, but existing systems often lack personalization, neglecting factors such as patient pain, participation, and recovery speed. Moreover, most digital platforms lack essential human support.
This study aimed to explore the engagement, safety, and clinical effectiveness of a personalized and adaptive app-based human-supported digital monitoring and rehabilitation program. Conducted as a prospective multi-center longitudinal cohort study, 127 patients were enrolled. A smart alert system managed undesired events, triggering physician interventions in case of suspected issues. The app collected data on dropout rates, complications, readmissions, Patient-Reported Outcome Measures (PROMS), and patient satisfaction.
Results showed a 2% readmission rate, with smart alerts potentially preventing 85% of flagged issues through timely doctor interventions. Program adherence reached 77%, and 89% of patients endorsed the program's use. The integration of personalized, human- backed digital solutions emerged as a transformative approach to standardize and enhance the rehabilitation journey post-TKA, while showing potential cost savings.

Abstract. Early identification and prediction of chronic pain in patients after total knee arthroplasty can significantly impact treatment strategies and improve patient satisfaction. This study introduces an innovative artificial intelligence model that predicts pain levels and pain evolution after TKA, empowering surgeons with insights for personalized patient care.
The pain intensity was measured with a visual analog scale on a mobile application, from 1650 knee arthroplasty patients from one week before surgery and up to 12 weeks after surgery. A training set was first used to identify patterns in the data that could best approximate pain trajectories. Out-of-sample pain trajectories were predicted by estimating pattern weights and reconstructing the remaining timepoints. Confidence intervals were calculated to determine prediction accuracy.
The model's accuracy was evaluated based on the percentage of predictions falling within 10 % of the true pain values. With an observation time of up to week 2, the model achieved 67% accuracy in forecasting pain levels for the next 4 weeks, and 61% accuracy for the next 10 weeks. By extending the observation time to week 4, the accuracy improved to 84% and 69% respectively.
The artificial intelligence model showed promising results in predicting pain evolution. By utilizing this model, surgeon teams can manage patient expectations and tailor pain management strategies. The model's predictions facilitate efficient tele- monitoring, enabling remote patient monitoring of patient with less good evolution prediction, reducing the need for frequent clinic visits. Incorporating this technology into surgical practice can enhance surgical outcomes and patient satisfaction.

Abstract. In the realm of Computer Assisted Surgery (CAS), the accurate localization of anatom- ical features and surgical tools is crucial for enhancing surgical outcomes. Important efforts have been recently focused toward markerless navigation to reduce intraoperative intrusiv- ity. Unfortunately, recent studies don’t always satisfy the required accuracy and precision to be used clinically. This research investigates the application of deep learning models for pose estimation of synthetic anatomical features and surgery instruments to improve the localization accuracy. Based on the models from CosyPose and Coupled-Iterative- Refinement methods, we applied a three-step pose estimation process. Using 3D meshes and BlenderProc we generated a synthetic RGB-D dataset of scenes including tibias, fe- murs, shoulder’s glenoids and surgical instruments (Cutting guide) to train and test our model. Moreover, we compared our results to the Point Pair Features (PPF) method, a conventional pose estimation algorithm based on point cloud data. We found a signifi- cant enhancement regarding the accuracy with our model, achieving sub-millimetric and sub-degree accuracy, surpassing the PPF algorithm. We also provided qualitative exam- ples of the estimated poses to visualize the accuracy of our model. While the proposed method shows promising results, challenges remain in particular passing from synthetic to real-world data. Future efforts will focus on collecting annotated real-world data.

Abstract. Continuum dexterous manipulators (CDMs) have shown great potential when inte- grated with computer assisted orthopaedic surgery (CAOS) systems for minimally invasive surgery (MIS). We hypothesize that the enhanced dexterity of CDMs may allow for greater access to target tissue through a single port when compared to traditional, rigid MIS in- struments. To assess such CDMs for intervertebral disc removal applications, a phantom study in the scope of MIS transforaminal lumbar interbody fusion (TLIF) was conducted to evaluate the achievable surgical workspace of the intervertebral disc (IVD) during disc space preparation. A CDM with 6 mm diameter and a conformable nitinol whisk tip was evaluated against three 135° lumbar curettes in a 2D L4-L5 IVD phantom by an ex- perienced spine surgeon. Improvements of up to 41.7% in reachable IVD workspace are achieved with the CDM, demonstrating its viability in improving outcomes for MIS spinal fusion.

Abstract. Computer-assisted surgery relies on precise labeling of patient anatomy using 3D images. Major part of this process is nowadays performed by deep-learning (DL) algorithms. However, the evaluation of automated segmentations using conventional metrics like Dice coefficient or Hausdorff distance has limitations, especially when assessing non-significant errors at the mesh level. To overcome this, we propose a novel metric (SSPC) focusing on significant surface disparities to enhance evaluation accuracy.

Abstract. Patient-specific computational models of the shoulder have the potential to further our understanding of joint biomechanics and optimise treatments for common pathologies such as osteoarthritis and rotator cuff tears. Since active motion and stability of the shoulder are mainly governed by the surrounding soft tissues, such models must be able to reliable and accurately predict muscle paths. Our aim was to develop and validate a computationally efficient line-segment muscle mapping model capable of mapping patient-specific muscle paths of the multi-pennate muscles of the deltoid and rotator cuff.

A triangular surface mesh was generated from segmentations of an anonymous male subject (75 years old, BMI 23) and muscle origin/insertion points were identified based upon previously reported data. A ‘convex hull’ algorithm identified optimal muscle fibre paths during 0-90° coronal plane abduction, sagittal plane flexion and axial rotation at neutral elevation. The model was capable of computing muscle lengths, moment arms and line of action for each muscle fibre.

The model had acceptable correlation with in vivo and cadaveric data from the literature. The model was also highly efficient, capable of mapping 42 muscle segments at 2.5° intervals of joint motion in less than 17 seconds.

The current model presents a patient specific method for modelling muscle multi-pennate muscles of the shoulder with high computational efficiency, requiring only the surface mesh inputs of the bony anatomy and muscle origin/insertion points without the need for commercial finite element contact detection software. The model may be used further for the study of shoulder musculoskeletal disorders.

Abstract. This paper introduces a novel in-silico platform for designing, optimizing and validating 3D ultrasound speckle-tracking algorithms that target the biomechanical characterization of knee collateral ligaments. The platform is based on a numerical model of a cadaveric knee following knee arthroplasty (TKA) with a posterior stabilized implant and with experimentally obtained subject-specific material properties of knee collateral ligaments. Applying the platform to develop a 3D ultrasound speckle-tracking algorithm shows promise as the resulting algorithm is capable to capture subtle deformations and strain patterns, with root mean square errors of 0.10 and 0.72 for maximal and minimal principal strains, respectively. However, optimization and in-silico validation are currently limited to one specimen and loading scenario, necessitating further research. Nevertheless, this work lays the foundation for advancing ultrasound-based biomechanical assessments from 2D to 3D, particularly in knee arthroplasty, with the potential for broader clinical impact on musculoskeletal health assessments.

Abstract. Introduction: Most of the literature on robotic-assisted total hip arthroplasty (THA) outcomes is derived from a single computerized tomography-based robotic (CT-RTHA) platform. The purpose of this study was to compare one-year patient reported outcome measures (PROMs) between a novel, fluoroscopy-based, robotic-assisted (FL-RTHA) system and a CT-RTHA system.

Materials and methods: A review of 85 consecutive FL-RTHA and 125 consecutive CT-RTHA was conducted. All cases were performed via a direct anterior approach by one of two surgeons, during the same time period, for a pre-operative diagnosis of osteoarthritis, avascular necrosis, or rheumatoid arthritis. Outcomes included one-year post-operative Veterans RAND-12 (VR-12) Physical (PCS)/Mental (MCS), Hip Disability and Osteoarthritis Outcome (HOOS) Pain/Physical Function (PS)/Joint Replacement (JR), and University of California Los Angeles (UCLA) Activity scores. The primary comparative endpoint was the magnitude of improvements between pre- and post-operative scores.

Results: Patients in the FL-RTHA cohort had lower pre-operative VR-12 PCS, HOOS Pain, HOOS-PS, HOOS-JR, and UCLA Activity scores compared to patients in the CT-RTHA cohort. Patients in the FL-RTHA cohort reported significantly greater improvements in HOOS-PS scores (-41.54 vs. -36.55; p=0.028) than patients in the CT-RTHA cohort. Both cohorts experienced similar rates of post-operative complications requiring reoperation/revision surgery (FL-RTHA 0% vs. CT-RTHA 3.20%; p=0.095).

Conclusions: Both robotic techniques produced similar excellent PROM scores at one-year post-operative. However, use of the novel, fluoroscopy-based robotic system resulted in greater improvements in HOOS-PS at one-year relative to the computerized tomography-based robotic technique.

Abstract. Robotic Ultrasound Systems are an emerging technology. Autonomous systems, which are not yet on the market, comprise detection of the scanning area, the ultrasound scan and image processing. However, contact gel application is performed manually. This work proposes a concept for an autonomous contact gel applicator to contribute to a full automation of robotic ultrasound scanning systems e.g. for 3D-imaging of the knee.

Abstract. Mobile C-arm x-ray machines are commonly used in orthopaedic trauma surgeries to visualize internal anatomy. However, the use of scouting images to aid C-arm positioning during these procedures can prolong operating time and increase radiation exposure. Our Depth Camera Augmented Fluoroscopy (DeCAF) device is designed to reduce the number of x-ray images needed by overlaying the fluoroscopic images onto a live video of the patient’s surface anatomy. In this study, we demonstrate in a simulated operating room (OR) environment that the DeCAF system has clinically acceptable overlay accuracy (1.3 ± 0.2 mm) and allowed the surgeon to substantially eliminate use of x-rays while fixing proximal tibial plates in acceptable positions without significantly changing the time required (p = 0.72). This justifies proceeding to live clinical evaluations.

Abstract. Computer Assisted Surgical (CAS) systems have been used successfully in joint arthroplasty to improve the accuracy of resections across multiple joint surgeries including hips, knees and shoulders. A Total Ankle Arthroplasty (TAA) application for CAS system was developed the intent of facilitating the procedure and improving accuracy of bone resections in the foot and ankle and evaluated for accuracy using artificial ankle joint specimens. The results demonstrate resections with errors of less than 2mm and 2° and therefore the system can offer both accurate and precise intraoperative surgical resection measurements during computer-assisted TAA.

Abstract. A Total Ankle Arthroplasty application for a Computer Assisted Surgery (CAS) system was developed the intent of facilitating the procedure and improving accuracy of bone resections in the foot and ankle. The application was evaluated for error sources using artificial ankle joint specimens. Mean and 95 % confidence intervals for positioning, execution and verification errors were less than 2mm and 2° for all specimens and each degree of freedom controlled by the CAS system.

Abstract. This study investigates quadriceps inhibition trends after total knee arthroplasty (TKA). Eighteen patients undertook perioperative physiotherapy using Sliderâ, a Class 1 medical device that measures forces and knee motion. Sliderâ monitors and uses gamification to encourage users to do knee exercises at home while displaying the results on their tablet. The results are sent securely to their clinician with exception notification. Perioperative changes in the straight leg raise (SLR), inner range quadriceps (IRQ), and knee flexion lying (KFL) exercises were analyzed. After starting Sliderâ two patients declined surgery because their knee pain had decreased, two moved away, two were cancelled for medical reasons, and twelve had TKA. Two patients developed infections and abnormal trends in the results of the knee exercises were detected by the Sliderâ algorithm before the patients’ clinic appointments. The curves of perioperative results of patients who had no infection were J-shaped (concave) postoperatively, while those of the patients with infections were m-shaped (double-convex). These objective findings are consistent with clinical experience of quadriceps inhibition occurring due to fluid accumulation following infection. In the infected patients, Sliderâ’s algorithm enabled timely alerts to the patients’ clinics and early intervention. One had a successful washout, the other needed a revision. Larger studies on Sliderâ are needed to provide statistically significant characterization of anomalies to support clinical decisions.

Abstract. 3D planning of a corrective osteotomy of a radius malunion requires a healthy reference which is not always available. A shape completion model could create a reference bone based on the unaffected part of the malunited bone. The aim of this study is to develop and validate a shape completion model for clinical use.
A statistical shape model (SSM) was developed based on CT scans of 80 healthy radii. This SSM was expanded into a shape completion model to predict the distal 12% of the radius, based on the proximal 88%. Nine other CT scans were used to validate this model and to set hyper parameters that dictate factors of the available data. Finally, eight more CT scans were used to test the performance of the shape completion model.
The average accuracy of the shape completion model, measured through a root mean square difference, was 0.42mm (SD 0.10) and the average Hausdorff distance was 3.57mm (SD 1.09). The predicted radii were comparable to the actual radii, with the differences mainly found at the radial styloid, Lister tubercle and the sigmoid notch. The radiocarpal articular surface was often well predicted.
In conclusion, the current shape completion model showed to predict clinically useable models that have error margins that are comparable with previous models described in literature. Future research should compare the use of the shape completion model to the use of the contralateral radius to find what technique works more optimal.

Abstract. Robotic-Assistance (RA) in total knee arthroplasty (TKA) has gained popularity due to its ability to improve accuracy and facilitate patient specific techniques. However, the reported influence on clinical outcomes is inconsistent. This study assessed the influence of the recently introduced VELYS Robotic-Assisted Solution on the Knee Society Function Score (KSFS) at early follow up. A retrospective review of a prospective company sponsored registry was conducted to assess KSFS at first post-operative follow up for all ATTUNE primary TKAs. RA-TKAs were differentiated from MI-TKA. The mean KSFS was compared between groups. Pre-operative demographic information including sex, age, BMI and pre-operative KSFS scores were compared and an ANCOVA was used on post-operative KSFS scores to account for any differences. There were 553 RA-TKAs and 6,710 MI-TKAs with KSFS reported at first post-operative follow-up. The proportion of females was similar between the groups but the RA-TKA group was older (69.37 Vs. 66.37, p<0.0001), had slightly lower BMI (29.90 Vs. 30.79, p<0.0001) and had post-operative follow up at an earlier time point (87.43 Vs. 116 days, p<0.0001). The mean post-operative KSFS was significantly higher for the RA-TKA group than the MI-TKA group (79.59 Vs. 76.44, p=0.0001 and 78.15 Vs. 74.05 following ANCOVA adjustment). In this study, KSFS was improved for RA compared to MI at first follow-up. These findings suggest that RA combined with patient specific techniques can improve patient outcomes. Further investigation with longer-term follow-up is required.

Abstract. Artificial intelligence (AI) and machine learning (ML) take an ever-growing place in medical care. Anatomical segmentation and reconstruction is one of the fields where ML reveals to be very efficient. Yet, verification of ML results still requires human verification and correction especially on pathologic morphologies. We propose an automatic assessment of AI-generated scapular reconstructions. Based on deep learning (DL), it separates predictions requiring little to no revision from predictions where corrected voxels represent more than 1% of the scapula, with an accuracy of 80%.

Abstract. Introduction 3D imaging technologies allow for more accurate planning of multiplanar corrective osteotomies around the knee. However, 3D planning introduces complexity. For example, varus/valgus is expressed in the coronal plane of the leg, while posterior tibial slope is expressed in the sagittal plane of the tibia. We present a software platform that deals with this complexity.
Methodology A high tibial osteotomy was parametrized using 5 parameters: hinge axis rotation in axial plane; hinge axis tilt in axial plane; hinge axis position along longitudinal axis; hinge axis distance to cortex; osteotomy opening angle. A MATLAB-algorithm was developed in which the effect of each possible combination of osteotomy parameters was obtained on change in alignment parameters expressed in coordinate systems of the tibia and leg. By finding the set of osteotomy parameters that approaches the predefined change in alignment parameters the closest, a 3D osteotomy planning can be linked to a desired outcome.
Results A digital interactive osteotomy planning platform was developed that automatically computes the required hinge axis orientation and osteotomy opening/closing angle for a given predefined outcome of change in posterior tibial slope and %-weight-bearing-line in the coordinate systems of the tibia and leg, respectively, with an error of <0.1°. During planning, the user is able to manually tweak the osteotomy and is given real-time feedback on the resulting trade-offs.
Conclusion The presented osteotomy planning platform allows for automatic and complex 3D planning of multiplanar leg malalignment correction, while integrating that alignment parameters are measured in different coordinate systems.

Abstract. In vitro physiological knee joint simulators have been proven to be valuable tools for characterizing knee joint biomechanics, complementing in vivo measurements. However, many simulators only allow simulations of squatting motions. This is partly due to the lack of a tailored approach to efficiently identify the required simulator input parameters for accurate investigation of other frequently performed activities of daily living (ADL). Therefore, we aimed to develop a novel in vivo-based computational approach which uniquely integrates multiple constraints from a novel in vitro knee simulator to determine the required muscle forces during various ADLs.
During a motion capture study, six healthy subjects performed squatting, sit-stand- sit and gait motions. Subject-scaled musculoskeletal models were adapted to include constraints of the knee joint simulator by including only quadriceps and hamstring muscle actuators, down-scaling ground reaction forces and applying constant hamstring force. Subsequently, muscle forces were computed for each motion using the Concurrent Optimization of Muscle Activations and Kinematics algorithm.
Afterwards, the in silico-based squatting results were retrospectively compared with previously performed in vitro experiments using the knee joint simulator, which actively controlled the quadriceps and bilateral hamstrings to maintain a constant vertical ground reaction force of 110N during squatting. Resulting in silico computed and simulator-measured forces during squatting showed similar magnitudes and high correlations. This indicates robustness of the proposed in vivo-based computational approach. Accordingly, its application to stance phase of gait initiated quasi-static in vitro simulations of this additional motion, further demonstrating its feasibility.

Abstract. In order to improve the usability of surgical navigation systems, this paper presents the design, implementation and testing of a mobile wireless tracking camera cart for intraoperative use. The tracking data, i.e., tool poses, are made available on the operating room (OR) network by means of the manufacturer-independent communication standard ISO IEEE 11073 (“SDC”). In this way, the wireless tracking system (WTS) can act as a modular localization service for various medical devices such as robots and navigation software.
The concept and implementation of the battery-powered wireless tracking cart is shown. The paper furthermore evaluates how the use of a new software library improves latency and reliability in contrast to previous implementations.
The results indicate that the new software library and the wireless data transfer do indeed meet the required latency of 50 milliseconds for hand-eye coordination tasks with a reliability of 99.395% even when the network is under OR-typical load. However, as the library does not offer any additional safety determinism, increased maximum latency times may occur in individual cases, making the device unsafe for intraoperative use.

Abstract. Many patients exhibit joint level biomechanics and mobility deficits after knee arthroplasty, which have been linked to patient dissatisfaction. Advancements in robotic-assisted surgery offer the potential for surgery personalization to address these deficits, yet the link between surgical planning and joint mechanics remains unclear. This research aims to comprehensively model the relationships among patient variability in joint mechanics, anatomy/morphology, physical activity, implant characteristics and post-operative outcomes to inform personalization strategies to address these deficits in joint mechanics. This is a five-year longitudinal patient cohort study with integrated data at several time points perioperatively including longitudinally during the pre-operative wait period, and post-operatively to one year. We combine patient-specific information from multiple domains including demographics and anthropometrics; patient-reported outcomes; three-dimensional gait kinematics through AI-driven markerless motion capture integrated into the clinic hallway; free-living physical activity (PA) and gait outcomes with inertial sensors; joint anatomy, morphology and OA feature modeling through custom CT image processing; and intraoperative robotic data. Longitudinal pre-operative outcomes have been collected for a subset (n=57) to date. There were no significant longitudinal changes in gait kinematics or objective PA outcomes on a population level pre-operatively. However, a subset of patients exhibited significant gait worsening, and worsening was significantly correlated with more advanced OA-related gait deficits at baseline. Unique free-living gait metrics were identified that moderately correlated with in-clinic gait kinematics.
PA outcomes were not significantly correlated to gait outcomes, and significantly worse PA outcomes (step count, %sedentary, %light and moderate to vigorous PA) were identified for female patients.

Abstract. Introduction
Joint-preserving tumor resection and patient-specific implant (PSI) reconstruction have the advantages of retaining the function and growth of native joints. Long-term complications of PSI in young active patients were a concern [1]. 3D planning and assistive tools like Navigation (NAVI) or 3D-printed guides (PSG) may facilitate the technically demanding surgical procedures [2,3] as it is challenging to translate the surgical plan with a negative surgical margin and match it to the PSI.

Methods
We treated 20 paediatric and adolescent patients with lower extremity bone sarcoma who underwent joint-preserving tumor resection and extendable PSI (2006 and 2019). The mean age was 11 (5-17). Engineers designed PSI and PSG based on surgeons' defined surgical planning in MIMICS (version 16, Materialise, Belgium). After neoadjuvant chemotherapy, surgeries were assisted with 12 NAVI (Stryker), 4 PSG, or a combination (4 NAVIG) (Figure 1).
Oncological results and limb functions were recorded. Patient overall disease survival and PSI survival with revision surgery as endpoint were calculated using Kaplan-Meier method (R Core Team (2023).

Results
With a mean follow-up of 9.9 years (4.7 -17.7), three patients died from disease progression, and one patient defaulted follow-up. Using Kaplan-Meier, patient survival was 90% and 81.8% at five and ten years (Figure 2A). PSI survival was 94.1% and 86.9% at five and ten years (Figure 2B). One Henderson Type II aseptic loosening and two Type III structural failures required major implant components to be revised. MSTS score was 29.2 out of 30 (28-30). Surgical margins were all negative in bone except one soft tissue positive margin. All PSIs matched well to the retained articular bones intraoperatively and showed osseointegration at PSI junctions (except one) and continuous joint growth postoperatively.

Discussion
With 3D planning and tools like NAVI and PSG, technically demanding joint-preserving surgeries could be accurately implemented as planned with excellent limb and joint function without compromising oncological results[4]. The computer-assisted approach might minimize soft tissue detachment by retaining the blood supply to the small articular bone segments. Osseointegration at the well-matched PSI junction suggested stable primary fixation and secondary healing of the PSI [5]. More cases and comparisons with conventional joint-sacrificing surgery were needed.

Abstract. Introduction
In orthopaedic oncology, computer navigation and 3D-printed guides facilitate precise osteotomies only after surgical exposure[1,2]. Mixed Reality is an immersive technology merging real and virtual worlds, and users can interact with digital objects[3]. Through Head-Mounted Displays, surgeons directly visualize holographic models that overlay tumor patients in their physical environment before surgeries start. Clinical reports of MR application are limited, and no data in orthopaedic oncology.

Methods
Between July 2021 and January 2024, we retrospectively reviewed 24 bone tumor patients undergoing surgeries. A holographic application was created using patients’ 2D medical images. In the conventional 2D method (Figure 1A), the surgeon studied 2D images and mentally overlaid the virtual 3D models onto the patients’ bodies. In the MR method (Figure 1B), the surgeons directly visualized 3D holograms on the patients’ bodies via HMD. Both methods were used to clinically assess the same patient. The surgeon completed 1) a Likert-Scale (LS) questionnaire to assess his opinions on the spatial awareness of the bone structures and the effectiveness of surgical planning and 2) The National Aeronautics and Space Administration-Task Load Index (NASA-TLX) score to evaluate the surgeons’ cognitive workload. The results of the two methods were compared using Wilcoxon Signed Rank Test.

Results
The Likert-scale questionnaire revealed that the 3D holograms in the MR technology group were more effective than the Conventional 2D group. For the cognitive workload for preoperative clinical assessment, the MR technology group received significantly lower “mental”, “performance” and “frustration” scores; however, they received significantly higher “physical demand” and “effort” ratings than the Conventional group.

Discussion and Conclusion
MR technology improved 3D visualization and spatial awareness of bone tumors in patients’ anatomies and may facilitate surgical planning before skin incisions in orthopaedic oncology surgery. The results concurred with the first case series of MR applications during orthopaedic surgery [4]. With less cognitive load and better ergonomics, surgeons can stay focused on the patients and surgical tasks while keeping their hands free and sterile to manipulate virtual objects [5]. Further studies can investigate whether MR technology guides and replicates surgical plans.

Abstract. Human Risk management poses a challenge for an open integration of medical devices in digital operating rooms according to the ISO IEEE 11073 SDC standard. A User Interface Profile has the potential to improve safety and usability, which defines the HMI characteristics of a medical device in an ensemble, aligning with ISO 14971 and IEC 62366-1 standards. We discuss applications in orthopedics, highlighting UI Profiles' potential to mitigate human-induced and process-related risks.