Radiological Imaging Evaluation of the Failing Total Hip Replacement

Diagnostic and prognostic relevance of plain radiographs for periprosthetic joint infections of the hip: a literature review

Conventional radiography is regularly used to evaluate complications after total hip arthroplasty. In various recent consensus meetings, however, plain radiographs of a potentially infected hip joint have been judged as being only relevant to exclude diagnoses other than infection. Solid data on radiographic presentations of periprosthetic joint infection (PJI) are scarce. As a result, the prognostic value of radiological features in low-grade PJI remains uncertain. The present review article aims to present an overview of the available literature and to develop ideas on future perspectives to define the diagnostic possibilities of radiography in PJIs of the hip. The primary outcome of interest of this systematic review was the radiologic presentation of periprosthetic joint infections of the hip. As secondary outcome of interest served the sensitivity and specificity of the radiologic presentation of periprosthetic joint infections. Of the included articles, 26 were reviews, essays, or case reports and only 18 were clinical studies. Typical radiologic abnormalities of PJI were a periosteal reaction, a wide band of radiolucency at the cement-bone or metal-bone interface, patchy osteolysis, implant loosening, bone resorption around the implant, and transcortical sinus tracts. The frequency of their occurrence is still inadequately defined. A deeper understanding of the underlying causes and the relation between microorganisms to radiologic abnormalities can probably help clinicians in the future to diagnose a PJI. This is why further research shall focus on the radiographic features of PJI.

[Tips and tricks of cement removal in the case of revision surgery]

BACKGROUND

In Germany, current revision rates after arthroplasty range between 28-37%. In particular, remaining cement residues are causative for additional revision surgery after periprosthetic joint infection, which is why complete cement removal is of high importance. However, the removal of remaining cement residues often confronts the surgeon with technical challenges. Complication-free and complete cement removal requires extensive preoperative preparation in order to develop the best possible surgical strategy.

TREATMENT

Special instrument sets to facilitate cement removal in revision cases are available from various manufacturers. In addition to endoluminal approaches, access enhancements such as extended osteotomies exist to facilitate complete cement removal. Finally, the surgeon should be able to give the indication for an intraoperative procedural change after a defined time interval.

Dynamic Acetabular Cup Orientation during Gait: A Study of Fast- and Slow-Walking Total Hip Replacement Patients

The dynamic orientation of total hip replacement acetabular cups during walking may vary substantially from their assumed position at surgical implantation and may vary between individuals. The scale of this effect is of interest for both pre-clinical device testing and for pre-operative surgical planning. This work aimed to evaluate (1) patient variation in dynamic cup orientation; (2) whether walking speed was a candidate proxy measure for the dynamic cup orientation; and (3) the relationships between dynamic cup orientation angles and planar pelvic angles. Pelvic movement data for patients with fast (20 patients) and slow (19 patients) self-selected walking speeds were used to calculate acetabular cup inclination and version angles through gait. For aim 1, the range and extremes of acetabular cup orientation angles were analysed for all patients. A large patient-to-patient variation was found in the ranges of both inclination angle (1° to 11°) and version angle (4° to 18°). The version angle was typically retroverted in comparison to the implantation position (greatest deviation 27°). This orientation is substantially different to the static, 0° version, simplifying assumptions in pre-clinical 'edge loading' testing. For aim 2, the cup orientation angles were compared between the fast- and slow-walking groups using statistical parametric mapping. The only significant differences observed were for cup version angle, during ~12% of the gait cycle before toe-off (p < 0.05). Therefore, self-selected walking speed, in isolation, is not a sufficient proxy measure for dynamic acetabular orientation. For aim 3, correlations were recorded between the acetabular cup orientation angles and the planar pelvic angles. The cup inclination angle during gait was strongly correlated (Spearman's coefficient -1) with pelvic obliquity alone, indicating that simple planar assessment could be used to anticipate inclination angle range. The cup version angle was correlated with both pelvic rotation and tilt (Spearman's coefficient 0.8-1), indicating that cup version cannot be predicted directly from any single pelvic movement. This complexity, along with the interaction between inclination angle and range of version angle, supports the use of computational tools to aid clinical understanding.

Periprosthetic femoral fractures

The incidence of periprosthetic femoral fractures is rising in the UK, because of an ageing population and an increasing number of hip arthroplasty operations being performed. They can occur intra- or postoperatively, and usually follow low energy trauma. They present with pain, swelling over the thigh, and an inability to weight bear. Periprosthetic femoral fractures are usually classified as per the unified classification system. Their management usually is dependent on their classification, with type A (fracture at level of greater or lesser trochanter) managed non-operatively with protected weight bearing, type B (fracture adjacent to implant) managed with either open reduction internal fixation or revision surgery, and type C (fracture distal to implant) managed with open reduction internal fixation. Owing to their complexity, these patients must be adequately optimised before surgery and appropriately rehabilitated.

AI-based hip prosthesis failure prediction through evolutional radiological indices

INTRODUCTION

This study aimed to develop artificial intelligence models for predicting hip implant failure from radiological features. Analyzing the evolution of the periprosthetic bone and implant's position throughout the entire follow-up period has shown the potential to be more relevant in outcome prediction than simply considering the latest radiographic images. Thus, we investigated an AI-based model employing a small set of evolutional parameters derived from conventional radiological features to predict hip prosthesis failure.

MATERIALS AND METHODS

One hundred sixty-nine radiological features were annotated from historical anteroposterior and lateral radiographs for 162 total hip arthroplasty patients, 32 of which later underwent implant failure. Linear regression on each patient's chronologically sorted radiological features was employed to derive 169 corresponding evolutional parameters per image. Three sets of machine learning predictors were developed: one employing the original features (standard model), one the evolutional ones (evolutional model), and the last their union (hybrid model). Each set included a model employing all the available features (full model) and a model employing the few most predictive ones according to Gini importance (minimal model).

RESULTS

The evolutional and hybrid predictors resulted highly effective (area under the ROC curve (AUC) of full models = 0.94), outperforming the standard one, whose AUC was only 0.82. The minimal hybrid model, employing just four features, three of which evolutional, scored an AUC of 0.95, proving even more accurate than the full one, exploiting 173 features. This tool could be shaped to be either highly specific (sensitivity: 80%, specificity: 98.6%) or highly sensitive (sensitivity: 90%, specificity: 92.4%).

CONCLUSION

The proposed predictor may represent a highly sensitive screening tool for clinicians, capable to predict THA failure with an advance between a few months and more than a year through only four radiological parameters, considering either their value at the latest visit or their evolution through time.

Lack of Consensus on the Definition of Aseptic Loosening in Total Ankle Replacement: A Narrative Systematic Review

Background: Aseptic loosening is one of the most common modes of failure of total ankle replacement (TAR). However, a precise definition of aseptic loosening is still lacking. This systematic review aimed to identify the variations of applied definitions and offer insights into the lack of consensus. Methods: Human studies reporting aseptic loosening of TAR published in peer-reviewed journals within the last decade were considered. The search strategy involved specific terms in Embase, MEDLINE ALL, and the Cochrane Library. Variations in aseptic loosening definitions were analysed. Results: Of 767 studies, 88 were included in this study. Only nine studies precisely defined aseptic loosening with significant variations. Twenty-two studies referenced the term and fifty-seven reported it as a complication but neither defined nor referenced it. Conclusions: Significant uncertainty exists regarding the universal definition of aseptic loosening of TAR, and many variations occur in terms of the assessment approach and criteria.

Cost-effectiveness of implant movement analysis in aseptic loosening after hip replacement: a health-economic model

OBJECTIVE

To investigate the cost-effectiveness of using Implant Movement Analysis (IMA) to follow up suspected aseptic loosening when the diagnosis after an initial X-ray is not conclusive, compared with a diagnostic pathway with X-ray follow-up.

METHODS

A health-economic model in the form of a decision tree was developed using quality-adjusted life years (QALY) from the literature, cost-per-patient data from a university hospital and the probabilities of different events from expert physicians' opinions. The base case incremental cost-effectiveness ratio (ICER) was compared with established willingness-to-pay thresholds and sensitivity analyses were performed to account for assumptions and uncertainty.

RESULTS

The base case ICER indicated that the IMA pathway was cost effective (SEK 99,681, compared with the SEK 500,000 threshold). In the sensitivity analysis, the IMA pathway remained cost effective during most changes in parameters. ICERs above the threshold value occurred in cases where a larger or smaller proportion of people receive immediate surgery.

CONCLUSION

A diagnostic pathway using IMA after an inconclusive X-ray for suspected aseptic loosening was cost effective compared with a pathway with X-ray follow-up.

Improved visualization of the bone-implant interface and osseointegration in ex vivo acetabular cup implants using photon-counting detector CT

BACKGROUND

Successful osseointegration of joint replacement implants is required for long-term implant survival. Accurate assessment of osseointegration could enable clinical discrimination of failed implants from other sources of pain avoiding unnecessary surgeries. Photon-counting detector computed tomography (PCD-CT) provides improvements in image resolution compared to conventional energy-integrating detector CT (EID-CT), possibly allowing better visualization of bone-implant-interfaces and osseointegration. The aim of this study was to assess the quality of visualization of bone-implant-interfaces and osseointegration in acetabular cup implants, using PCD-CT compared with EID-CT.

METHODS

Two acetabular implants (one cemented, one uncemented) retrieved during revision surgery were scanned using PCD-CT and EID-CT at equal radiation dose. Images were reconstructed using different reconstruction kernels and iterative strengths. Delineation of the bone-implant and bone-cement-interface as an indicator of osseointegration was scored subjectively for image quality by four radiologists on a Likert scale and assessed quantitatively.

RESULTS

Delineation of bone-implant and bone-cement-interfaces was better with PCD-CT compared with EID-CT (p ≤ 0.030). The highest ratings were given for PCD-CT at sharper kernels for the cemented cup (PCD-CT, median 5, interquartile range 4.25-5.00 versus EID-CT, 3, 2.00-3.75, p < 0.001) and the uncemented cup (5, 4.00-5.00 versus 2, 2-2, respectively, p < 0.001). The bone-implant-interface was 35-42% sharper and the bone-cement-interface was 28-43% sharper with PCD-CT compared with EID-CT, depending on the reconstruction kernel.

CONCLUSIONS

PCD-CT might enable a more accurate assessment of osseointegration of orthopedic joint replacement implants.

KEY POINTS

• The bone-implant interface ex vivo showed superior visualization using photon-counting detector computed tomography (PCD-CT) compared to energy-integrating detector computed tomography. • Harder reconstruction kernels in PCD-CT provide sharper images with lower noise levels. • These improvements in imaging might make it possible to visualize osseointegration in vivo.

Influence of Scan Parameters of Single and Dual-Energy CT Protocols in Combination with Metal Artifact Suppression Algorithms for THA: An ex Vivo Study

BACKGROUND

Metal artifacts caused by hip arthroplasty stems limit the diagnostic value of computed tomography (CT) in the evaluation of periprosthetic fractures or implant loosening. The aim of this ex vivo study was to evaluate the influence of different scan parameters and metal artifact algorithms on image quality in the presence of hip stems.

METHODS

Nine femoral stems, 6 uncemented and 3 cemented, that had been implanted in subjects during their lifetimes were exarticulated and investigated after death and anatomical body donation. Twelve CT protocols consisting of single-energy (SE) and single-source consecutive dual-energy (DE) scans with and without an iterative metal artifact reduction algorithm (iMAR; Siemens Healthineers) and/or monoenergetic reconstructions were compared. Streak and blooming artifacts as well as subjective image quality were evaluated for each protocol.

RESULTS

Metal artifact reduction with iMAR significantly reduced the streak artifacts in all investigated protocols (p = 0.001 to 0.01). The best subjective image quality was observed for the SE protocol with a tin filter and iMAR. The least streak artifacts were observed for monoenergetic reconstructions of 110, 160, and 190 keV with iMAR (standard deviation of the Hounsfield units: 151.1, 143.7, 144.4) as well as the SE protocol with a tin filter and iMAR (163.5). The smallest virtual growth was seen for the SE with a tin filter and without iMAR (4.40 mm) and the monoenergetic reconstruction of 190 keV without iMAR (4.67 mm).

CONCLUSIONS

This study strongly suggests that metal artifact reduction algorithms (e.g., iMAR) should be used in clinical practice for imaging of the bone-implant interface of prostheses with either an uncemented or cemented femoral stem. Among the iMAR protocols, the SE protocol with 140 kV and a tin filter produced the best subjective image quality. Furthermore, this protocol and DE monoenergetic reconstructions of 160 and 190 keV with iMAR achieved the lowest levels of streak and blooming artifacts.

LEVEL OF EVIDENCE

Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.

[Strategies for stem revision : Surgery planning, implant removal and reimplantation]

BACKGROUND AND PLANNING

Femoral revision represents a challenging procedure in its surgical planning and technical execution. The results are mainly dependent on the bone loss present, the quality of the implant removal and the choice of revision system. Patient-specific factors such as age, comorbidities, bone quality or even the localization of the cement have a decisive influence on the surgical strategy. Appropriate preoperative preparation, including the presence of the necessary special instruments, which are essential for gentle implant removal, is indispensable to achieve the best possible result.

THERAPY

A bone-sparing implant removal can be particularly challenging, especially in the case of well-fixed stems and cement residues, which may extend far beyond the isthmus. In such situations, a transfemoral approach should be considered. Cementless reimplantation using modular or non-modular titanium stems is currently the preferred treatment of choice for most revisions. In elderly, low-demand patients or with poor bone quality, cemented revision remains a good alternative treatment option.

Influence of Metal Implants on Quantitative Evaluation of Bone Single-Photon Emission Computed Tomography/Computed Tomography

When visualizing biological activity at nonunion sites by the radioisotopes, gamma rays are more attenuated if metal implants are placed in the bone. However, the effects of various implant types and their placement on gamma ray attenuation in quantitative evaluation remain unknown. To elucidate these effects, we created a phantom that simulated the nonunion of the femur in this study. The count of gamma rays was measured by single-photon emission computed tomography/computed tomography (SPECT/CT) while considering CT-based attenuation correction (CTAC), metal implant placement, type (intramedullary nail or plate), and position. The count differed significantly with and without CTAC and with and without implants (both types) under CTAC. Significantly different counts were observed between the intramedullary nail and plate placed contralaterally to the lesion (i.e., non-lesion side). No significant difference was observed between the intramedullary nail and plate on the lesion side or between plates on the non-lesion and lesion sides. The measured standardized uptake value (SUV) was closer to the true SUV with CTAC than without. Moreover, the count was higher with implants than without. However, even with implants, it was lower than the actual count, indicating the absence of overcorrection. Implant type and position do not seem to influence the count.

Application of 68Ga-citrate PET/CT for differentiating periprosthetic joint infection from aseptic loosening after joint replacement surgery

AIMS

We aimed to evaluate the utility of ⁶⁸Ga-citrate positron emission tomography (PET)/CT in the differentiation of periprosthetic joint infection (PJI) and aseptic loosening (AL), and compare it with ^99mTc-methylene bisphosphonates (^99mTc-MDP) bone scan.

METHODS

We studied 39 patients with suspected PJI or AL. These patients underwent ⁶⁸Ga-citrate PET/CT, ^99mTc-MDP three-phase bone scan and single-photon emission CT (SPECT)/CT. PET/CT was performed at ten minutes and 60 minutes after injection, respectively. Images were evaluated by three nuclear medicine doctors based on: 1) visual analysis of the three methods based on tracer uptake model, and PET images attenuation-corrected with CT and those not attenuation-corrected with CT were analyzed, respectively; and 2) semi-quantitative analysis of PET/CT: maximum standardized uptake value (SUVmax) of lesions, SUVmax of the lesion/SUVmean of the normal bone, and SUVmax of the lesion/SUVmean of the normal muscle. The final diagnosis was based on the clinical and intraoperative findings, and histopathological and microbiological examinations.

RESULTS

Overall, 23 and 16 patients were diagnosed with PJI and AL, respectively. The sensitivity and specificity of three-phase bone scan and SPECT/CT were 100% and 62.5%, 82.6%, and 100%, respectively. Attenuation correction (AC) at 60 minutes and non-AC at 60 minutes of PET/CT had the same highest sensitivity and specificity (91.3% and 100%), and AC at 60 minutes combined with SPECT/CT could improve the diagnostic efficiency (sensitivity = 95.7%). Diagnostic efficacy of the SUVmax was low (area under the curve (AUC) of ten minutes and 60 minutes was 0.814 and 0.806, respectively), and SUVmax of the lesion/SUVmean of the normal bone at 60 minutes was the best semi-quantitative parameter (AUC = 0.969).

CONCLUSION

⁶⁸Ga-citrate showed the potential to differentiate PJI from AL, and visual analysis based on uptake pattern of tracer was reliable. The visual analysis method of AC at 60 minutes, combined with ^99mTc-MDP SPECT/CT, could improve the sensitivity from 91.3% to 95.7%. In addition, a major limitation of our study was that it had a limited sample size, and more detailed studies with a larger sample size are warranted. Cite this article: Bone Joint Res 2022;11(6):398-408.

Postoperative Fluid Collections in Total Joint Arthroplasty: A Narrative Review

A post-operative fluid collection (POFC) represents a common finding in both primary and revision total joint arthroplasty (TJA). Fortunately, most resolve on their own, but in instances where they become symptomatic, prompt identification and management are paramount, especially when they occur adjacent to a joint arthroplasty because of the increased the risk of developing a periprosthetic joint infection. A strong clinical suspicion with appropriate clinical exam is required along with select imaging modalities to arrive at a diagnosis. Meticulous surgical technique is crucial to prevent POFC, but new emerging treatments continue to evolve. This article presents an updated overview of incidence, pathophysiology, diagnosis, and management of POFC in the setting of TJA. We review the role of select imaging modalities as well as summarize current literature regarding new treatments such as sclerotherapy agents, acellular dermal matrices, and negative pressure wound therapy. Future studies are necessary to explore the interplay of inflammatory mediators in POFC formation and to define their role in fluid collection resolution.

5-Year Clinical and Radiographic Results of the Direct Anterior Approach for Total Hip Arthroplasty Using a Collared Cementless Femoral Short-Stem Prosthesis

The aim of this study was to investigate the radiological and clinical outcome of the direct anterior approach (DAA) in total hip arthroplasty (THA) using a collared cementless femoral short-stem. This retrospective study included 124 patients with 135 THAs operated from 2014 to 2016 using a collared cementless triple tapered hydroxyapatite-coated femoral short-stem (AMIStem H Collared^®, Medacta International, Castel San Pietro, Switzerland) implanted with a DAA. Follow-up was performed at three months, 12 months, and five years. Clinical outcome was assessed using the hip osteoarthritis outcome score (HOOS) and radiological analysis was done using conventional radiographs, which included evaluation of the femur morphology based on Dorr classification, of radiolucencies based on the Gruen zone classification and of stem subsidence. The mean age was 67.7 ± 11.3 years and the mean body mass index (BMI) was 27.4 ± 4.4 kg/m². The stem survival rate at five years was 99.1% with one revision due to recurrent dislocations. Mean HOOS score improved from 40.9 ± 18.3 preoperatively to 81.5 ± 19.7 at three months, 89.3 ± 10.9 at 12 months, and 89.0 ± 14.0 at five years (all with p < 0.001). No significant correlations were found between age, femoral bone morphology, BMI and HOOS, and the appearance of relevant radiolucencies.

Role of 99mTc-Methylene Diphosphonate Bone Scintigraphy in the Evaluation of the Viability and the Incorporation of the Bone Allograft Used in Orthopedic Reconstruction

Allogenic bone is the most commonly grafted tissue which provides only osteoconductive property, in which it acts as a scaffold to facilitate the ingrowth of the vessels and migration of host cells capable of osteogenesis. It can be used as a substitute for autografts as the latter is associated with morbidity and limited donor site availability. Its applications are expanding in all aspects of orthopedic surgery, notably in revision hip replacement or surgical treatment for bone tumors or benign conditions. The balance between osteolysis and osteogenesis must be maintained for graft incorporation to occur and thus postoperative imaging is essential for differentiation between grafts and recurrent disease or viability/nonviability. Here, we present three cases, in which bone allografting was done who underwent serial 99 mTc-methylene diphosphonate three-phase bone scintigraphy with single-photon emission computed tomography/computed tomography to assess the viability, integrity, and the incorporation of the graft.

SPECT-CT metabolic and morphological study of 2 types of cemented hip stem prostheses in primary total hip arthroplasty patients: A protocol for a randomized controlled clinical trial (SPECT-PROTMA)

BACKGROUND

Cemented hip arthroplasty requires applying a layer of polymethylmethacrylate (cement) in the space between the bone and the prosthetic stem. This can be achieved using 2 techniques: the thick-layer technique (requires a layer of at least 2 mm to surround an undersized prosthetic stem), and the thin-layer technique (requires a thin layer of cement, so that the prosthetic stem fills the femoral medullary canal). Both approaches have excellent long-term clinical and radiological outcomes, although an implant's insertion into the bone generates inevitable bone mass and bone metabolic changes around it. Combination of single photon emission computed tomography and computed tomography scan (SPECT-CT) imaging combines the single photon emission computed tomography's ability to provide detailed bone metabolism assessment with the computed tomography scan's capacity to provide a meticulous anatomical study.

METHODS

This is a single center, open label, randomized clinical trial, performed in the premises of the Bellvitge University Hospital. Participants will be randomly assigned to the Thick-layer technique group (Exeter V40 Cemented Femoral Stem) or to the French paradox technique group (Müller Straight Stem). All participants will have a SPECT-CT scan study at 3, 6, 12, and 24 months after the surgery.

DISCUSSION

Surgical distress itself and the implant's insertion into the bone may cause microvascular changes that alter periprosthetic bone mass and bone metabolism. To the best of our knowledge, there are no studies using SPECT-CT to compare bone metabolism evolution in the postoperative period between these 2 surgical cementation techniques. We aim to provide information in this regard that could help decision making in complicated implant cases and, maybe, pave the way for larger, and methodologically improved studies.

TRIAL REGISTRATION

NCT05010733 (August 18, 2021).

Updates in postoperative imaging modalities following musculoskeletal surgery

Postoperative imaging following orthopaedic surgeries is essential in assessing complications post-surgery and also helps plan further treatment. Combining a high degree of clinical insight with appropriate imaging can guide the treating clinician to the correct diagnosis. Imaging is quite challenging because of surgery-related soft tissue changes, especially in the early postoperative period and the presence of metal implants resulting in image scatter and metal artifacts. Newer modalities and advances in imaging have helped overcome shortcomings and assess better, especially in procedures that involve implants. Collaborative decision-making involving radiologists and clinicians has shown to be beneficial and is the way forward. This narrative review discusses the utility of imaging in evaluating postoperative complications following musculoskeletal surgeries with specific relation to trauma, arthroplasty, and tumour by discussing commonly encountered clinical scenarios.

Accuracy and reproducibility of visual estimation of the acetabular cup positioning in total hip arthroplasty on plain radiographs by orthopaedic surgeons

OBJECTIVES

To examine whether trauma and orthopaedic surgeons could visually assess the anteversion and inclination of a total hip replacement acetabular component from standard anteroposterior radiograph and anteversion on a standard lateral radiograph with accuracy or reproducibility.Main outcome measurement: The main outcome was accuracy of visual estimations of angles. The secondary outcome was whether these estimations were reproducible though intra-observer variability.

RESULTS

Mean angles of anteversion on the anteroposterior, inclination on the anteroposterior and anteversion on the lateral on formal measurements were 15.2°, 45.4° and 19.9°; and the visual estimates were 17.5°, 45.9° and 18.2°, respectively. When comparing the visual estimates of surgeons and formal measurements, the results ranged from very poor to very good. Intra-observer reproducibility was moderate for all angles. The difference between the consultants and speciality registrars was not significant.

CONCLUSION

This study illustrated that not all orthopaedic surgeons were able to visually estimate angles well. Although some of our participants were very accurate, there were some who statistically were very poor. This level of inaccuracy can lead to inconsistency and we strongly suggested specialist software is used to assess acetabular cup position on postoperative plane radiographs rather than relying on 'visual estimations'.

Methods to address metal artifacts in post-processed CT images - A do-it-yourself guide for orthopedic surgeons

Computed tomography (CT) scans are often used for postoperative imaging in orthopedics. In the presence of metallic hardware, artifacts are generated, which can hamper visualization of the CT images, and also render the study ineffective for 3-D printing. Various solutions are available to minimize metal artifacts, and radiologists can employ these before or after processing the CT study. However, the orthopedic surgeon may be faced with situations where the metal artifacts were not addressed. To counter such problems, we present three do-it-yourself (DIY) techniques that can be used to manage metal artifacts.

Deep Learning Artificial Intelligence Model for Assessment of Hip Dislocation Risk Following Primary Total Hip Arthroplasty From Postoperative Radiographs

BACKGROUND

Dislocation is a common complication following total hip arthroplasty (THA), and accounts for a high percentage of subsequent revisions. The purpose of this study is to illustrate the potential of a convolutional neural network model to assess the risk of hip dislocation based on postoperative anteroposterior pelvis radiographs.

METHODS

We retrospectively evaluated radiographs for a cohort of 13,970 primary THAs with 374 dislocations over 5 years of follow-up. Overall, 1490 radiographs from dislocated and 91,094 from non-dislocated THAs were included in the analysis. A convolutional neural network object detection model (YOLO-V3) was trained to crop the images by centering on the femoral head. A ResNet18 classifier was trained to predict subsequent hip dislocation from the cropped imaging. The ResNet18 classifier was initialized with ImageNet weights and trained using FastAI (V1.0) running on PyTorch. The training was run for 15 epochs using 10-fold cross validation, data oversampling, and augmentation.

RESULTS

The hip dislocation classifier achieved the following mean performance (standard deviation): accuracy = 49.5 (4.1%), sensitivity = 89.0 (2.2%), specificity = 48.8 (4.2%), positive predictive value = 3.3 (0.3%), negative predictive value = 99.5 (0.1%), and area under the receiver operating characteristic curve = 76.7 (3.6%). Saliency maps demonstrated that the model placed the greatest emphasis on the femoral head and acetabular component.

CONCLUSION

Existing prediction methods fail to identify patients at high risk of dislocation following THA. Our radiographic classifier model has high sensitivity and negative predictive value, and can be combined with clinical risk factor information for rapid assessment of risk for dislocation following THA. The model further suggests radiographic locations which may be important in understanding the etiology of prosthesis dislocation. Importantly, our model is an illustration of the potential of automated imaging artificial intelligence models in orthopedics.

LEVEL OF EVIDENCE

Level III.

Acetabular and Femoral Component Positioning Using Direct Anterior Approach Versus Posterior Approach in Total Hip Arthroplasty

PURPOSE

Direct anterior approach (DAA) has recently become popular in total hip arthroplasty (THA). However, irrespective of the surgical approach used, component malposition is an important factor affecting function and complications after THA. This study aims to compare component positioning on the femoral and acetabular side between DAA and posterior approach (PA) to the hip joint. We hypothesized that the two approaches are similar in terms of component positioning.

METHODS

We prospectively studied 50 patients, matched according to age, sex, and body mass index, undergoing THA, divided non-randomly into 2 groups. Group 1 comprised 25 patients (35 hips) undergoing THA using DAA and group 2 comprised 25 patients (25 hips) undergoing THA using PA. Ten patients from group 1 had simultaneous bilateral THA. Radiological parameters studied were acetabular inclination (AI), coronal femoral stem alignment (CFA), leg length difference (LLD), acetabular cup version (AV), and femoral stem version (FV).

RESULTS

There was no significant difference in AI, CFA, LLD, AV, and FV between the two groups. Excellent to good inter and intra-observer reliability expressed in terms of intraclass correlation coefficient (ICC) was noted for all the radiographic measurements.

CONCLUSION

Both DAA and PA for THA achieve comparable radiological component positioning. DAA may not provide any advantage over PA in terms of positioning of the prosthesis.

LEVEL OF EVIDENCE

Level II, non-randomized comparative study.

Artificial intelligence in radiotherapy

Artificial intelligence (AI) has already been implemented widely in the medical field in the recent years. This paper first reviews the background of AI and radiotherapy. Then it explores the basic concepts of different AI algorithms and machine learning methods, such as neural networks, that are available to us today and how they are being implemented in radiotherapy and diagnostic processes, such as medical imaging, treatment planning, patient simulation, quality assurance and radiation dose delivery. It also explores the ongoing research on AI methods that are to be implemented in radiotherapy in the future. The review shows very promising progress and future for AI to be widely used in various areas of radiotherapy. However, basing on various concerns such as availability and security of using big data, and further work on polishing and testing AI algorithms, it is found that we may not ready to use AI primarily in radiotherapy at the moment.