Multicentric development and validation of a multi-scale and multi-task deep learning model for comprehensive lower extremity alignment analysis.

Osteoarthritis of the knee, a widespread cause of knee disability, is commonly treated in orthopedics due to its rising prevalence. Lower extremity misalignment, pivotal in knee injury etiology and management, necessitates comprehensive mechanical alignment evaluation via frequently-requested weight-bearing long leg radiographs (LLR). Despite LLR's routine use, current analysis techniques are error-prone and time-consuming. To address this, we conducted a multicentric study to develop and validate a deep learning (DL) model for fully automated leg alignment assessment on anterior-posterior LLR, targeting enhanced reliability and efficiency. The DL model, developed using 594 patients' LLR and a 60%/10%/30% data split for training, validation, and testing, executed alignment analyses via a multi-step process, employing a detection network and nine specialized networks. It was designed to assess all vital anatomical and mechanical parameters for standard clinical leg deformity analysis and preoperative planning. Accuracy, reliability, and assessment duration were compared with three specialized orthopedic surgeons across two distinct institutional datasets (136 and 143 radiographs). The algorithm exhibited equivalent performance to the surgeons in terms of alignment accuracy (DL: 0.21 ± 0.18°to 1.06 ± 1.3°vs. OS: 0.21 ± 0.16°to 1.72 ± 1.96°), interrater reliability (ICC DL: 0.90 ± 0.05 to 1.0 ± 0.0 vs. ICC OS: 0.90 ± 0.03 to 1.0 ± 0.0), and clinically acceptable accuracy (DL: 53.9%-100% vs OS 30.8%-100%). Further, automated analysis significantly reduced analysis time compared to manual annotation (DL: 22 ± 0.6 s vs. OS; 101.7 ± 7 s, p ≤ 0.01). By demonstrating that our algorithm not only matches the precision of expert surgeons but also significantly outpaces them in both speed and consistency of measurements, our research underscores a pivotal advancement in harnessing AI to enhance clinical efficiency and decision-making in orthopaedics.

Significant difference in femoral torsion between coronal plane alignment of the knee type 1 and 4.

PURPOSE: The purpose of this study was to find out whether the torsions of the femur and tibia are dependent on the coronal plane alignment of the knee (CPAK) type. METHODS: Five hundred patients (1000 legs) were included, who received a whole leg standing three-dimensional (3D) radiograph using EOS imaging (EOS Imaging, Paris, France). SterEOS software was used for digital reconstruction. Femoral and tibial torsions were determined by analysing 3D reconstructions of each leg. Femoral torsion was defined as the angle between the femoral neck axis (FNA) and the posterior condylar axis (PCA). Tibial torsion was defined as the angle between the axis tangent to the posterior part of the tibia plateau and the transmalleolar axis. Arithmetic hip-knee-ankle angle (aHKA) and joint-line obliquity (JLO) were also determined, allowing each leg to be assigned one of nine possible phenotypes according to CPAK. RESULTS: The mean femoral torsion in CPAK type 1 was significantly higher (+ 2.6° ± 0.8°) than in CPAK type 4 (p = 0.02). All other CPAK types did not differ in the degree of femoral torsions. No differences could be demonstrated for the tibial torsion. CONCLUSION: There is a correlation between the coronal alignment of the lower limb and femoral torsion. This may provide the basis for extending the CPAK classification beyond the coronal plane. LEVEL OF EVIDENCE: Level III.

[Isolated Patellofemoral Osteoarthritis]. / Isolierte Patellofemoralarthrose.

Isolated patellofemoral osteoarthritis is a complex pathology. It occurs frequently in women over 40 years and leads to a high degree of suffering. The causes of femoropatellar degeneration are manifold and require stage-specific therapy. This is very challenging for the treating physician. This article is intended to provide a structured overview of the symptoms, diagnostics and stage-specific therapy of patellofemoral osteoarthritis.

Definition of normal, neutral, deviant and aberrant coronal knee alignment for total knee arthroplasty.

PURPOSE: One of the most pertinent questions in total knee arthroplasty (TKA) is: what could be considered normal coronal alignment? This study aims to define normal, neutral, deviant and aberrant coronal alignment using large data from a computed tomography (CT)-scan database and previously published phenotypes. METHODS: Coronal alignment parameters from 11,191 knee osteoarthritis (OA) patients were measured based on three dimensional reconstructed CT data using a validated planning software. Based on these measurements, patients' coronal alignment was phenotyped according to the functional knee phenotype concept. These phenotypes represent an alignment variation of the overall hip knee ankle angle (HKA), femoral mechanical angle (FMA) and tibial mechanical angle (TMA). Each phenotype is defined by a specific mean and covers a range of ±1.5° from this mean. Coronal alignment is classified as normal, neutral, deviant and aberrant based on distribution frequency. Mean values and distribution among the phenotypes are presented and compared between two populations (OA patients in this study and non-OA patients from a previously published study). RESULTS: The arithmetic HKA (aHKA), combined normalised data of FMA and TMA, showed that 36.0% of knees were neutral within ±1 SD from the mean in both angles, 44.3% had either a TMA or a FMA within ±1-2 SD (normally aligned), 15.3% of the patients were deviant within ±2-3 SD and only 4.4% of them had an aberrant alignment (±3-4 SD in 3.4% and >4 SD in 1.0% of the patients respectively). However, combining the normalised data of HKA, FMA and TMA, 15.4% of patients were neutral in all three angles, 39.7% were at least normal, 27.7% had at least one deviant angle and 17.2% had at least one aberrant angle. For HKA, the males exhibited 1° varus and females were neutral. For FMA, the females exhibited 0.7° more valgus in mean than males and grew 1.8° per category (males grew 2.1° per category). For TMA, the males exhibited 1.3° more varus than females and both grew 2.3° and 2.4° (females) per category. Normal coronal alignment was 179.2° ± 2.8-5.6° (males) and 180.5 > ± 2.8-5.6° (females) for HKA, 93.1 > ± 2.1-4.2° (males) and 93.8 > ± 1.8-3.6° (females) for FMA and 86.7 > ± 2.3-4.6° (males) and 88 > ± 2.4-4.8° (females) for TMA. This means HKA 6.4 varus or 4.8° valgus (males) or 5.1° varus to 6.1° valgus was considered normal. For FMA HKA 1.1 varus or 7.3° valgus (males) or 0.2° valgus to 7.4° valgus was considered normal. For TMA HKA 7.9 varus or 1.3° valgus (males) or 6.8° varus to 2.8° valgus was considered normal. Aberrant coronal alignment started from 179.2° ± 8.4° (males) and 180.5 > ± 8.4° (females) for HKA, 93.1 > ± 6.3° (males) 93.8 > ± 5.4° (females) for FMA and 86.7 > ± 6.9° (males) and 88 > ± 7.2° (females) for TMA. This means HKA > 9.2° varus or 7.6° valgus (males) or 7.9° varus to 8.9° valgus was considered aberrant. CONCLUSION: Definitions of neutrality, normality, deviance as well as aberrance for coronal alignment in TKA were proposed in this study according to their distribution frequencies. This can be seen as an important first step towards a safe transition from the conventional one-size-fits-all to a more personalised coronal alignment target. There should be further definitions combining bony alignment, joint surfaces' morphology, soft tissue laxities and joint kinematics. LEVEL OF EVIDENCE: III.

Insights into the Distribution Patterns of Foot and Ankle Tumours: Update on the Perspective of a University Tumour Institute.

The rarity of foot and ankle tumours, together with the numerous histological entities, presents a challenge in accumulating sufficient patients to draw reliable conclusions. Therefore, we decided to present an update of a retrospective analysis of their distribution patterns, comprising 536 cases of foot and ankle tumours presented to our tumour board between June 1997 and June 2023. Our aim was to provide a comprehensive overview of the prevalence and distribution patterns of benign and malignant bone and soft tissue tumours of the foot and ankle. A total of 277 tumours involved bone (51.7%). Of these, 242 (87.4%) were benign and 35 (12.6%) were malignant. In addition, 259 soft tissue tumours (48.3%) were found, of which 191 (73.7%) were benign and 68 (26.3%) were malignant. The most common benign bone tumours were simple bone cysts, enchondromas, osteochondromas, aneurysmal bone cysts, and lipomas of bone. Common benign soft tissue tumours included a tenosynovial giant cell tumour, haemangioma, plantar fibromatosis, schwannoma, and lipoma. The most common malignant soft tissue tumours were synovial sarcoma, malignant melanoma, and myxofibrosarcoma. In terms of anatomical location, the hindfoot was the most common site (28.7%), followed by the midfoot (25.9%), ankle (25.4%), and forefoot (20.0%). The distribution of benign entities often follows typical patterns, which may facilitate an early diagnosis even without biopsy (e.g., simple bone cyst, plantar fibromatosis). On the other hand, the distribution patterns of many rare or malignant entities are inconsistent. Individual soft tissue malignancies occur very sporadically, even over long periods of time and in specialized tumour centres. It is therefore important to recognise that any suspicious mass in the foot and ankle must be considered a possible malignancy until proven otherwise.

Goal-directed Perioperative Albumin Substitution Versus Standard of Care to Reduce Postoperative Complications: A Randomized Clinical Trial (SuperAdd Trial).

OBJECTIVE: To investigate whether goal-directed albumin substitution during surgery and postanesthesia care to maintain a serum albumin concentration >30 g/L can reduce postoperative complications. BACKGROUND: Hypoalbuminemia is associated with numerous postoperative complications. Since albumin has important physiological functions, substitution of patients with hypoalbuminemia is worth considering. METHODS: We conducted a single-center, randomized, controlled, outcome assessor-blinded clinical trial in adult patients, American Society of Anesthesiologists physical status classification 3 to 4 or undergoing high-risk surgery. Patients, whose serum albumin concentration dropped <30 g/L were randomly assigned to goal-directed albumin substitution maintaining serum concentration >30 g/L or to standard care until discharge from the postanesthesia intermediate care unit. Standard of care allowed albumin substitution in hemodynamic instable patients with serum concentration <20 g/L, only. Primary outcome was the incidence of postoperative complications ≥2 according to the Clavien-Dindo Classification in at least 1 of 9 domains (pulmonary, infectious, cardiovascular, neurological, renal, gastrointestinal, wound, pain, and hematological) until postoperative day 15. RESULTS: Of 2509 included patients, 600 (23.9%) developed serum albumin concentrations <30 g/L. Human albumin 60 g (40-80 g) was substituted to 299 (99.7%) patients in the intervention group and to 54 (18.0%) in the standard care group. At least 1 postoperative complication classified as Clavien-Dindo Classification ≥2 occurred in 254 of 300 patients (84.7%) in the intervention group and in 262 of 300 (87.3%) in the standard treatment group (risk difference -2.7%, 95% CI, -8.3% to 2.9%). CONCLUSION: Maintaining serum albumin concentration of >30 g/L perioperatively cannot generally be recommended in high-risk noncardiac surgery patients.

A New Method for Assessing Patients' Obesity-Associated Infection Risk Using X-rays in Hip Arthroplasties.

Overweight patients have higher complication rates during and after surgical procedures. In total hip arthroplasty (THA), postoperative infection is a major complication. In this study, we show that the patient's body mass index (BMI) can be approximated by a newly developed grading system using preoperative X-rays. Furthermore, we show that a higher score and BMI result in a higher risk of infection. For this retrospective study, 635 patients undergoing THA or revision surgeries in 2018 and 2019 were included. The preoperatively acquired X-rays of the pelvis were analyzed using a four-stage grading system. The infection rate was compared to our score and the patients' BMI. The mean BMI (95% confidence) of all patients graded as grade 0 was 25.16 (24.83; 25.50) kg/m2, for grade 1, it was 30.31 (29.52; 31.09) kg/m2, for grade 2, it was 35.06 (33.59; 36.54) kg/m2, and it was 45.03 (39.65; 50.41) kg/m2 for grade 3. The risk of infection was 4% in patients with normal radiographs, rising from 7% in patients graded as 1 up to 18% in each of the highest categories. This study shows that we were able to create a semi-quantitative grading tool for the abdominal contour displayed on X-rays of the pelvis in order to estimate the patients' BMI and therefore the infection rate. A higher abdominal contour grade showed higher infection rates at follow-up.

Deep Learning-Based Detection of Bone Tumors around the Knee in X-rays of Children.

Even though tumors in children are rare, they cause the second most deaths under the age of 18 years. More often than in other age groups, underage patients suffer from malignancies of the bones, and these mostly occur in the area around the knee. One problem in the treatment is the early detection of bone tumors, especially on X-rays. The rarity and non-specific clinical symptoms further prolong the time to diagnosis. Nevertheless, an early diagnosis is crucial and can facilitate the treatment and therefore improve the prognosis of affected children. A new approach to evaluating X-ray images using artificial intelligence may facilitate the detection of suspicious lesions and, hence, accelerate the referral to a specialized center. We implemented a Vision Transformer model for image classification of healthy and pathological X-rays. To tackle the limited amount of data, we used a pretrained model and implemented extensive data augmentation. Discrete parameters were described by incidence and percentage ratio and continuous parameters by median, standard deviation and variance. For the evaluation of the model accuracy, sensitivity and specificity were computed. The two-entity classification of the healthy control group and the pathological group resulted in a cross-validated accuracy of 89.1%, a sensitivity of 82.2% and a specificity of 93.2% for test groups. Grad-CAMs were created to ensure the plausibility of the predictions. The proposed approach, using state-of-the-art deep learning methodology to detect bone tumors on knee X-rays of children has achieved very good results. With further improvement of the algorithm, enlargement of the dataset and removal of potential biases, this could become a useful additional tool, especially to support general practitioners for early, accurate and specific diagnosis of bone lesions in young patients.

Accuracy of Positioning and Risk Factors for Malpositioning Custom-Made Femoral Stems in Total Hip Arthroplasty-A Retrospective Multicenter Analysis.

Total hip arthroplasty (THA) is commonly performed using off-the-shelf implants. In the case of a severe mismatch between the anatomy of the proximal femur and the geometry of the stem, the use of custom-made stems might become necessary. The goal of this study was to investigate the precision of the implantation of custom-made stems of one manufacturer (CTX stem, AQ Implants) and to determine risk factors for malpositioning. All patients receiving a custom-made CTX stem between 2014 and 2020 at six high-volume academic centers were retrospectively recruited. The achieved position of the stem, as determined by stem version, stem coronal angle, and implantation depth on radiographs, was compared to the plan. The influence of radiographic and demographic parameters on the position was investigated. The results revealed a high variability of the achieved implant position in relation to the preoperative plan. While the stem coronal angle only differed slightly from the intended position, the stem version and the implantation depth showed a high frequency and amount of deviation. Right stems showed significantly higher positions than planned. Surgeons must be aware of this potential problem when implanting custom-made stems.

Synovial Calprotectin for Diagnosing Periprosthetic Joint Infection in Loose Hip and Knee Arthroplasties: A Prospective Cohort Study.

BACKGROUND/AIM: Synovial calprotectin has been demonstrated as a promising biomarker for periprosthetic joint infections (PJI) in painful total hip (THA) and knee arthroplasties (TKA). However, its diagnostic utility has not been evaluated explicitly in cases with marked loosening or migration of the implant. Concerns have already been raised in cases with metallosis and severe periprosthetic osteolysis because wear-induced inflammation may yield false positive results. The purpose of this study was to evaluate calprotectin for the diagnosis of PJI in cases that preoperatively demonstrate moderate to severe periprosthetic osteolysis or implant migration as signs for implant loosening in THA and TKA. PATIENTS AND METHODS: Thirty-three patients were included in this prospective study between February of 2019 and November of 2021. The extent of osteolysis was classified according to Engh et al., Paprosky et al., and the modern Knee Society Radiographic Evaluation and Scoring System. Synovial white blood cell count (WBC), percentage of polymorphonuclear neutrophils (PMC), serum C-reactive protein (CRP) and synovial calprotectin using a lateral-flow-assay were tested against the European Bone and Joint Infection Society (EBJIS) definition for PJI. Statistic quality criteria were calculated and compared using a binary classification test. RESULTS: Ten patients were classified as confirmed infections according to the EBJIS definition (7 THA and 5 TKA). The calprotectin assay yielded a sensitivity of 0.60, a specificity of 0.61, a positive predictive value of 0.40, and a negative predictive value of 0.78. The calprotectin assay resulted in nine false positive and four false negative cases. No correlation between the extent of osteolysis and false classification by means of the calprotectin assay was observed. CONCLUSION: The diagnostic accuracy of synovial calprotectin is impaired if moderate to severe signs of implant loosening are present. If PJI is unlikely, the calprotectin LFT can be applied as a further exclusion tool as the negative predictive value remains relatively high.

"Koehlers teardrop is not a reliable landmark for assessing the centre of rotation after Total hip arthroplasty" - a retrospective radiological study.

PURPOSE: Various anatomical landmarks have become established in radiography for the assessment of cup positioning after total hip arthroplasty (THA). The most important one is Koehler's teardrop figure (KTF). However, there is a lack of data on the validity of this landmark, which is widely used clinically for assessing the centre of rotation of the hip. METHOD: A retrospective measurement of the lateral and cranial distance of the KTF to the centre of hip rotation was performed on the basis of 250 X-ray images of patients who had undergone THA. In addition, the dependence of these distances on pelvic tilt was determined in 16 patients by means of virtual X-ray projections based on pelvic CTs. RESULTS: It was shown that the distance of the KTF from the centre of hip rotation in the horizontal plane is gender-dependent (men: 42.8 ± 6.0 mm vs. women: 37.4 ± 4.7 mm; p < 0.001) and age-dependent (Pearson correlation - 0.114; p < 0.05). Furthermore, the vertical and horizontal distances are subject to variation depending on height (Pearson correlation 0.14; p < 0.05 and 0.40; p < 0.001, respectively) and weight (Pearson correlation 0.158; p < 0.05). The distance between the KTF and the centre of hip rotation varies slightly depending on pelvic tilt. CONCLUSION: The KTF is not a sufficiently valid landmark for assessing the centre of rotation after THA. It is influenced by many different disturbance variables. However, it is largely robust against changes in pelvic tilt, so that it can be used as a reference point when comparing different intraindividual radiographs to assess the change in the centre of rotation due to implantation or to detect cup migration.

Postoperative full leg radiographs exhibit less residual coronal varus deformity compared to intraoperative measurements in robotic arm-assisted total knee arthroplasty with the MAKO™ system.

PURPOSE: Robotic arm-assisted total knee arthroplasty (raTKA), currently a major trend in knee arthroplasty, aims to improve the accuracy of implant positioning and limb alignment. However, it is unclear whether and to what extent manual radiographic and navigation measurements with the MAKO™ system correlate. Nonetheless, a high agreement would be crucial to reliably achieve the desired limb alignment. METHODS: Thirty-six consecutive patients with osteoarthritis and a slight-to-moderate varus deformity undergoing raTKA were prospectively included in this study. Prior to surgery and at follow-up, a full leg radiograph (FLR) under weight-bearing conditions was performed. In addition, a computed tomography (CT) scan was conducted for preoperative planning. The hip-knee-ankle angle (HKA), mechanical lateral distal femur angle (mLDFA), mechanical medial proximal tibial angle (mMPTA) and joint line convergence angle (JLCA) were measured in the preoperative and follow-up FLR as well as in the CT scout (without weight-bearing) by three independent raters. Furthermore, the HKA was intraoperatively assessed with the MAKO™ system before and after raTKA. RESULTS: Significantly higher HKA values were identified for intraoperative deformity assessment using the MAKO system compared to the preoperative FLR and CT scouts (p = 0.006; p = 0.05). Intraoperative assessment of the HKA with final implants showed a mean residual varus deformity of 3.2° ± 1.9°, whereas a significantly lower residual varus deformity of 1.4° ± 1.9° was identified in the postoperative FLR (p < 0.001). The mMPTA was significantly higher in the preoperative FLR than in the CT scouts (p < 0.001). Intraoperatively, the mMPTA was adjusted to a mean of 87.5° ± 0.9° with final implants, while significantly higher values were measured in postoperative FLRs (p < 0.001). Concerning the mLDFA, no significant differences could be identified. CONCLUSION: The clinical importance of this study lies in the finding that there is a difference between residual varus deformity measured intraoperatively with the MAKO™ system and those measured in postoperative FLRs. This has implications for preoperative planning as well as intraoperative fine-tuning of the implant position during raTKA to avoid overcorrection of knees with slight-to-moderate varus osteoarthritis. LEVEL OF EVIDENCE: Level IV.

Digital TKA Alignment Training with a New Digital Simulation Tool (Knee-CAT) Improves Process Quality, Efficiency, and Confidence.

Individual alignment techniques have been introduced to restore patients' unique anatomical variations during total knee arthroplasty. The transition from conventional mechanical alignment to individualised approaches, with the assistance of computer and/or robotic technologies, is challenging. The objective of this study was to develop a digital training platform with real patient data to educate and simulate various modern alignment philosophies. The aim was to evaluate the training effect of the tool by measuring the process quality and efficiency, as well as the post-training surgeon's confidence with new alignment philosophies. Based on 1000 data sets, a web-based interactive TKA computer navigation simulator (Knee-CAT) was developed. Quantitative decisions on bone cuts were linked to the extension and flexion gap values. Eleven different alignment workflows were introduced. A fully automatic evaluation system for each workflow, with a comparison function for all workflows, was implemented to increase the learning effect. The results of 40 surgeons with different experience levels using the platform were assessed. Initial data were analysed regarding process quality and efficiency and compared after two training courses. Process quality measured by the percentage of correct decisions was increased by the two training courses from 45% to 87.5%. The main reasons for failure were wrong decisions on the joint line, tibia slope, femoral rotation, and gap balancing. Efficiency was obtained with a reduction in time spent per exercise from 4 min 28 s to 2 min 35 s (42%) after the training courses. All volunteers rated the training tool as helpful or extremely helpful for learning new alignment philosophies. Separating the learning experience from OR performance was mentioned as one of the main advantages. A novel digital simulation tool for the case-based learning of various alignment philosophies in TKA surgery was developed and introduced. The simulation tool, together with the training courses, improved surgeon confidence and their ability to learn new alignment techniques in a stress-free out-of-theatre environment and to become more time efficient in making correct alignment decisions.

The impact of different alignment strategies on bone cuts in total knee arthroplasty for varus knee phenotypes.

PURPOSE: The purpose of this study was to visualise the influence of alignment strategy on bone resection in varus knee phenotypes. The hypothesis was that different amounts of bone resection would be required depending on the alignment strategy chosen. Through visualisation of the corresponding bone sections, it was hypothesised, it would be possible to assess which of the different alignment strategies would require the least amount of change to the soft tissues for the chosen phenotype, whilst still ensuring acceptable alignment of the components, and thus could be considered the most ideal alignment strategy. METHODS: Simulations of the different alignment strategies (mechanical, anatomical, constrained kinematic and unconstrained kinematic) in relation to their bone resections were performed on five common exemplary varus knee phenotypes. VARHKA174° VARFMA87° VARTMA84°, VARHKA174° VARFMA90° NEUTMA87°, VARHKA174° NEUFMA93° VARTMA84°, VARHKA177° NEUFMA93° NEUTMA87° and VARHKA177° VALFMA96° VARTMA81°. The phenotype system used categorises knees based on overall limb alignment (i.e. hip knee angle) but also takes into account joint line obliquity (i.e. TKA and FMA) and has been applied in the global orthopaedic community since its introduction in 2019. The simulations are based on long-leg radiographs under load. It is assumed that a change of 1° in the alignment of the joint line corresponds to a displacement of the distal condyle by 1 mm. RESULTS: In the most common phenotype VARHKA174° NEUFMA93° VARTMA84°, a mechanical alignment would result in an asymmetric elevation of the tibial medial joint line by 6 mm and a lateral distalisation of the femoral condyle by 3 mm, an anatomical alignment only by 0 and 3 mm, a restricted by 3 and 3 mm, respectively, whilst a kinematic alignment would result in no change in joint line obliquity. In the similarly common phenotype 2 VARHKA174° VARFMA90° NEUTMA87° with the same HKA, the changes are considerably less with only 3 mm asymmetric height change on one joint side, respectively, and no change in restricted or kinematic alignment. CONCLUSION: This study shows that significantly different amounts of bone resection are required depending on the varus phenotype and the alignment strategy chosen. Based on the simulations performed, it can, therefore, be assumed that an individual decision for the respective phenotype is more important than the dogmatically correct alignment strategy. By including such simulations, the modern orthopaedic surgeon can now avoid biomechanically inferior alignments and still obtain the most natural possible knee alignment for the patient.

The synovial fluid calprotectin lateral flow test for the diagnosis of chronic prosthetic joint infection in failed primary and revision total hip and knee arthroplasty.

PURPOSE: The diagnostic criteria of prosthetic joint infection (PJI) recommended by the most commonly used diagnostic algorithms can be obscured or distorted by other inflammatory processes or aseptic pathology. Furthermore, the most reliable diagnostic criteria are garnered during revision surgery. A robust, reliable addition to the preoperative diagnostic cascade is warranted. Calprotectin has been shown to be an excellent diagnostic marker for PJI. In this study, we aimed to evaluate a lateral flow test (LFT) in the challenging patient cohort of a national referral centre for revision arthroplasty. METHODS: Beginning in March 2019, we prospectively included patients scheduled for arthroplasty exchange of a total hip (THA) or knee arthroplasty (TKA). Synovial fluid samples were collected intra-operatively. We used the International Consensus Meeting of 2018 (ICM) score as the gold standard. We then compared the pre-operative ICM score with the LFT result to calculate its diagnostic accuracy as a standalone pre-operative marker and in combination with the ICM score as part of an expanded diagnostic workup. RESULTS: A total of 137 patients with a mean age of 67 (± 13) years with 53 THA and 84 TKA were included. Ninety-nine patients (72.8%) were not infected, 34 (25.0) were infected, and four (2.9%) had an inconclusive final score and could not be classified after surgery. The calprotectin LFT had a sensitivity (95% confidence interval) of 0.94 (0.80-0.99) and a specificity of 0.87 (0.79-0.93). The area under the receiver operating characteristic curve (AUC) for the calprotectin LFT was 0.94 (0.89-0.99). In nine cases with an inconclusive pre-operative ICM score, the calprotectin LFT would have led to the correct diagnosis of PJI. CONCLUSIONS: The synovial fluid calprotectin LFT shows excellent diagnostic metrics both as a rule-in and a rule-out test, even in a challenging patient cohort with cases of severe osteolysis, wear disease, numerous preceding surgeries, and poor soft tissue conditions, which can impair the common diagnostic criteria. As it is available pre-operatively, this test might prove to be a very useful addition to the diagnostic algorithm.

Topical vancomycin powder does not affect patella cartilage degeneration in primary total knee arthroplasty and conversion rate for secondary patella resurfacing.

INTRODUCTION: Vancomycin powder (VP) is an antibiotic first introduced in pediatric spinal surgery to prevent surgical site infections (SSI). Recently its topical application was expanded to total hip and knee arthroplasty (THA, TKA) and anterior cruciate ligament reconstruction (ACLR). Toxicity to cartilage is the subject of current research. The aim of this study was to prove the hypothesis that topical application of VP in TKA does not result in a degeneration of patella cartilage. We propagate that the conversion rate for secondary patella resurfacing is not influenced by its use. MATERIALS AND METHODS: Between 2014 and 2021, 4292 joints were included in this monocentric retrospective cohort study. All patients underwent TKA without primary patella resurfacing. After a change of the procedure in the hospital, one group (VPG) was administered VP intraoperatively. The other group (nVPG) received no VP during surgery (nVPG). The remaining perioperative procedure was constant over the investigation period. Conversion rates for secondary patella resurfacing for both groups were determined without making distinctions in the indication. A second cohort was composed of patients presenting for follow-up examination 12 months after TKA and included 210 joints. Retrospective radiographic evaluations were performed preoperatively, before discharge and at follow-up examination. Patella axial radiographs were analyzed for patella tracking (lateral patellar tilt, patellar displacement) and patella degeneration (Sperner classification, patellofemoral joint space). RESULTS: There was no significant difference in the conversion rate for secondary patella resurfacing (4.24% VPG, 4.97% nVPG). Patella tracking and patella degeneration did not differ significantly between both groups. CONCLUSIONS: The topical application of VP does not influence the conversion rate for secondary patella resurfacing. Moreover, it does not result in a degeneration of patella cartilage in TK. LEVEL OF EVIDENCE: Retrospective case series, Level III.

SAM-X: sorting algorithm for musculoskeletal x-ray radiography.

OBJECTIVE: To develop a two-phased deep learning sorting algorithm for post-X-ray image acquisition in order to facilitate large musculoskeletal image datasets according to their anatomical entity. METHODS: In total, 42,608 unstructured and pseudonymized radiographs were retrieved from the PACS of a musculoskeletal tumor center. In phase 1, imaging data were sorted into 1000 clusters by a self-supervised model. A human-in-the-loop radiologist assigned weak, semantic labels to all clusters and clusters with the same label were merged. Three hundred thirty-two non-musculoskeletal clusters were discarded. In phase 2, the initial model was modified by "injecting" the identified labels into the self-supervised model to train a classifier. To provide statistical significance, data split and cross-validation were applied. The hold-out test set consisted of 50% external data. To gain insight into the model's predictions, Grad-CAMs were calculated. RESULTS: The self-supervised clustering resulted in a high normalized mutual information of 0.930. The expert radiologist identified 28 musculoskeletal clusters. The modified model achieved a classification accuracy of 96.2% and 96.6% for validation and hold-out test data for predicting the top class, respectively. When considering the top two predicted class labels, an accuracy of 99.7% and 99.6% was accomplished. Grad-CAMs as well as final cluster results underlined the robustness of the proposed method by showing that it focused on similar image regions a human would have considered for categorizing images. CONCLUSION: For efficient dataset building, we propose an accurate deep learning sorting algorithm for classifying radiographs according to their anatomical entity in the assessment of musculoskeletal diseases. KEY POINTS: • Classification of large radiograph datasets according to their anatomical entity. • Paramount importance of structuring vast amounts of retrospective data for modern deep learning applications. • Optimization of the radiological workflow and increase in efficiency as well as decrease of time-consuming tasks for radiologists through deep learning.

The impact of different alignment strategies on bone cuts for neutral knee phenotypes in total knee arthroplasty.

PURPOSE: The purpose of this study was to simulate and visualise the influence of the alignment strategy on bone resection in neutral knee phenotypes. It was hypothesised that different amounts of bone resection would be required depending on the alignment strategy chosen. The hypothesis was that by visualising the corresponding bone cuts, it would be possible to assess which of the different alignment strategies required the least change to the soft tissues for the chosen phenotype but still ensured acceptable component alignment and could, therefore, be considered the most ideal alignment strategy. METHODS: Simulations of the different alignment strategies (mechanical, anatomical, restricted kinematic and unrestricted kinematic) regarding their bone resections were performed on four common exemplary neutral knee phenotypes. NEUHKA0° VARFMA 90° VALTMA90°, NEUHKA0° NEUFMA 93° NEUTMA87°, NEUHKA0° VALFMA 96° NEUTMA87° and NEUHKA0° VALFMA 99° VARTMA84°. The phenotype system used categorises knees based on overall limb alignment (i.e. hip knee angle) but also considers joint line obliquity (i.e. TKA and FMA) and has been used globally since its introduction in 2019. These simulations are based on long leg weightbearing radiographs. It is assumed that a change of 1° in the alignment of the joint line corresponds to correspond to 1 mm of distal condyle offset. RESULTS: In the most common neutral phenotype NEUHKA0° NEUFMA 93° NEUTMA87°, with a prevalence of 30%, bone cuts remain below 4 mm regardless of alignment strategy. The greatest changes in the obliquity of the joint line can be expected for the mechanical alignment of the phenotype NEUHKA0° VALFMA 99° VARTMA84° where the medial tibia is raised by 6 mm and the lateral femur is shifted distally by 9 mm. In contrast, the NEUHKA0° VARFMA 90° VALTMA90° phenotype requires no change in joint line obliquity if the mechanical alignment strategy is used. CONCLUSION: Illustrations of alignment strategies help the treating surgeon to estimate the postoperative joint line obliquity. When considering the alignment strategy, it seems reasonable to prefer a strategy where the joint line obliquity is changed as little as possible. Although for the most common neutral knee phenotype the choice of alignment strategy seems to be of negligible importance, in general, even for neutral phenotypes, large differences in bone cuts can be observed depending on the choice of alignment strategy.