Factors related to disagreement in implant size between preoperative CT-based planning and the actual implants used intraoperatively for total hip arthroplasty

Two-stage multi-task deep learning framework for simultaneous pelvic bone segmentation and landmark detection from CT images

PURPOSE

Pelvic bone segmentation and landmark definition from computed tomography (CT) images are prerequisite steps for the preoperative planning of total hip arthroplasty. In clinical applications, the diseased pelvic anatomy usually degrades the accuracies of bone segmentation and landmark detection, leading to improper surgery planning and potential operative complications.

METHODS

This work proposes a two-stage multi-task algorithm to improve the accuracy of pelvic bone segmentation and landmark detection, especially for the diseased cases. The two-stage framework uses a coarse-to-fine strategy which first conducts global-scale bone segmentation and landmark detection and then focuses on the important local region to further refine the accuracy. For the global stage, a dual-task network is designed to share the common features between the segmentation and detection tasks, so that the two tasks mutually reinforce each other's performance. For the local-scale segmentation, an edge-enhanced dual-task network is designed for simultaneous bone segmentation and edge detection, leading to the more accurate delineation of the acetabulum boundary.

RESULTS

This method was evaluated via threefold cross-validation based on 81 CT images (including 31 diseased and 50 healthy cases). The first stage achieved DSC scores of 0.94, 0.97, and 0.97 for the sacrum, left and right hips, respectively, and an average distance error of 3.24 mm for the bone landmarks. The second stage further improved the DSC of the acetabulum by 5.42%, and this accuracy outperforms the state-of-the-arts (SOTA) methods by 0.63%. Our method also accurately segmented the diseased acetabulum boundaries. The entire workflow took ~ 10 s, which was only half of the U-Net run time.

CONCLUSION

Using the multi-task networks and the coarse-to-fine strategy, this method achieved more accurate bone segmentation and landmark detection than the SOTA method, especially for diseased hip images. Our work contributes to accurate and rapid design of acetabular cup prostheses.

The role of 3-dimensional preoperative planning for primary total hip arthroplasty based on artificial intelligence technology to different surgeons: A retrospective cohort study

Preoperative planning with computed tomography (CT)-based 3-dimensiona (3D) templating has been achieved precise placement of hip components. This study investigated the role of the software (3-dimensional preoperative planning for primary total hip arthroplasty [THA] based on artificial intelligence technology, artificial intelligence hip [AIHIP]) for surgeons with different experience levels in primary THA. In this retrospective cohort study, we included patients, who had undergone THA with the help of the AIHIP, and matched to patients, who had undergone THA without the help of the AIHIP, by age and the doctor who operated on them. The subjects were divided into 4 groups, senior surgeon (Chief of Surgery) with AIHIP group, senior surgeon without AIHIP group, junior surgeon (Associate Chief of Surgery) with AIHIP group and junior surgeon without AIHIP group. The general data, imaging index, clinical outcomes and accuracy of stem size prediction and cup size prediction were retrospectively documented for all patients. There was a significant difference in discrepancy in leg length (P = .010), neck-shaft angle (P = .025) and femoral offset (P = .031) between the healthy side and the affected side, operation duration (P < .001), decrease in hemoglobin (Hb) per 24 hours (P = .046), intraoperative radiation exposure frequency (P < .050) and postoperative complications (overall P = .035) among the patients in junior surgeon group. No significant differences were found between senior surgeon groups with respect to discrepancy in leg length (P = .793), neck-shaft angle (P = .088)and femoral offset (P = .946) between the healthy side and the affected side, operation duration (P = .085), decrease in Hb per 24 hours (P = .952), intraoperative radiation exposure frequency (P = .094) and postoperative complications (overall P = .378). The stem sizes of 95% were accurately estimated to be within 1 stem size, and 97% of the cup size estimates were accurate to within 1 cup size in senior surgeon group with AIHIP. A total of 87% stem sizes were accurately estimated to be within 1 stem size, and 85% cup sizes were accurate to within 1 cup size in junior surgeon group with AIHIP. In conclusion, our study suggests that an AI-based preoperative 3D planning system for THA is a valuable adjunctive tool for junior doctor and should routinely be performed preoperatively.

Acetabular Diameter Assessment and Three-Dimensional Simulation for Acetabular Reconstruction in Dysplastic Hips

BACKGROUND

This study aimed to investigate the relationship between acetabular width, three-dimensional (3D) simulation, and surgical results in total hip arthroplasty patients who have developmental dysplasia of the hip (DDH).

METHODS

This retrospective study included 216 DDH cases. Inner and outer acetabular width (OAW) was measured at the plane passing through the center of acetabular fossa. 3D simulation and 2D standard templating were performed. The actual cup size and the use of augments during surgery were recorded. Association among the indices and their distribution in different types of DDH were analyzed.

RESULTS

A difference of 13 to 14 millimeters (mm) was found between the inner acetabular width and actual cup size used in type II, III, and IV cases, while the difference was 0.2 to 3.6 mm for OAW. The accuracy of 2D templating and 3D simulation in predicting cup size was comparable in Crowe type I (86.5 versus 76%, P = .075), type II (72.7 versus 51.5%, P = .127), and type III (93.3 versus 66.7%, P = .169). The 3D simulation was significantly more accurate in Crowe type IV (89.1% versus 60.9%, P = .001). Augments and bone grafts were significantly more commonly used in type II (25%) than in the other types (0 to 6.5%).

CONCLUSION

OAW more accurately predicted actual cup size than inner acetabular width. The supero-lateral acetabular bone defects in type II cases require additional attention. Compared with 2D templating, 3D simulation is more accurate in predicting actual cup size in dysplastic hips with severe deformity and may be recommended in these selected cases, especially for Crowe IV patients.

Validation of CT-Based Three-Dimensional Preoperative Planning in Comparison with Acetate Templating for Primary Total Hip Arthroplasty

OBJECTIVE

This study aims to compare the accuracy of CT-based preoperative planning with that of acetate templating in predicting implant size, neck length, and neck cut length, and to evaluate the reproducibility of the two methods.

METHODS

This prospective study was conducted between August 2020 and March 2021. Patients who underwent elective primary total hip arthroplasty by a single surgeon were assessed for eligibility. The included patients underwent both acetate templating and CT-based planning by two observers after the operation. Each observer conducted both acetate templating and CT-based planning twice for each case. The outcome measures included the following: (1) the accuracy of surgical planning in predicting implant size, calcar length, and neck length, which was defined as the difference between the planned size and length and the actual size and length; (2) reproducibility of the two planning techniques, which were assessed by inter-observer and intra-observer reliability analysis; (3) the influence of potential confounding factors on planning accuracy, which was evaluated using generalized estimating equations.

RESULTS

A total of 57 cases were included in the study. CT-based planning was more accurate than acetate templating for predicting cup size (93% vs 79%, p < 0.001) and stem size (93% vs 75%, p < 0.001). When assessed by mean absolute difference, the comparison between acetate templating and CT-based planning was 4.28 mm vs 3.74 mm (p = 0.122) in predicting neck length and 3.05 mm vs 2.93 mm (p = 0.731) in predicting neck cut length. In the inter-observer reliability analysis, an intraclass correlation coefficient (ICC) of 0.790 was achieved for predicting cup size, and an ICC of 0.966 was achieved for predicting stem size using CT-based planning. In terms of intra-observer reliability, Observer 1 achieved an ICC of 0.803 for predicting cup size and 0.965 for predicting stem size in CT-based planning. Observer 2 achieved ICC values of 0.727 and 0.959 for predicting cup and stem sizes, respectively. The average planning time was 6.48 ± 1.55 min for CT-based planning and 6.12 ± 1.40 min for acetate templating (p = 0.015).

CONCLUSION

The CT-based planning system is more accurate than acetate templating for predicting implant size and has good reproducibility in total hip arthroplasty.

Predicting Implant Size in Total Hip Arthroplasty

Background

Efficient resource management is becoming more important as the demand for total hip arthroplasty (THA) increases. The purpose of this study is to evaluate the ability of linear regression and Bayesian statistics in predicting implant size for THA using patient demographic variables.

Material and methods

A retrospective, single-institution joint-replacement registry review was performed on patients who underwent primary THA from 2005 to 2019. Demographic information was obtained along with primary THA implant data. A total of 11,730 acetabular and 8536 femoral components were included. A multivariable regression model was created on a training cohort of 80% of the sample and applied to the validation cohort (remaining 20%). Bayesian posterior probability methods were applied to the training cohort and then tested in the validation cohort to determine the 1%, 5%, and 10% error tolerance thresholds.

Results

The most predictive regression model included height, weight, and sex (cup: R² = 0.57, all P < .001; stem mediolateral size [M/L]: R² = 0.32, all P < .001). Removing weight had a minimal effect and resulted in a more parsimonious model (cup: R² = 0.56, all P < .001; stem M/L: R² = 0.32, all P < .001). Applying the posterior probability estimate to the validation cohort in the Bayesian model using height, weight, and sex demonstrated high accuracy in predicting the range of required implant sizes (95.3% cup and 90.4% stem M/L size).

Conclusion

Implant size in THA is correlated with demographic variables to accurately predict implant size using Bayesian modeling. Predictive models such as linear regression and Bayesian modeling can be used to improve operating room efficiency, supply chain inventory management, and decrease costs associated with THA.

Utility of a novel integrated deep convolutional neural network for the segmentation of hip joint from computed tomography images in the preoperative planning of total hip arthroplasty

Purpose

Preoperative three-dimensional planning is important for total hip arthroplasty. To simulate the placement of joint implants on computed tomography (CT), pelvis and femur must be segmented. Accurate and rapid segmentation of the hip joint is challenging. This study aimed to develop a novel deep learning network, named Changmugu Net (CMG Net), which could achieve accurate segmentation of the femur and pelvis.

Methods

The overall deep neural network architecture of CMG Net employed three interrelated modules. CMG Net included the 2D U-net to separate the bony and soft tissues. The modular hierarchy method was used for the main femur segmentation to achieve better performance. A layer classifier was adopted to localise femur layers among a series of CT scan images. The first module was a modified 2D U-net, which separated bony and soft tissues; it provided intermediate supervision for the main femur segmentation. The second module was the main femur segmentation, which was used to distinguish the femur from the acetabulum. The third module was the layer classifier, which served as a post-processor for the second module.

Results

There was a much greater overlap in accuracy results with the “gold standard” segmentation than with competing networks. The dice overlap coefficient was 93.55% ± 5.57%; the mean surface distance was 1.34 ± 0.24 mm, and the Hausdorff distance was 4.19 ± 1.04 mm in the normal and diseased hips, which indicated greater accuracy than the other four competing networks. Moreover, the mean segmentation time of CMG Net was 25.87 ± 2.73 s, which was shorter than the times of the other four networks.

Conclusions

The prominent segmentation accuracy and run-time of CMG Net suggest that it is a reliable method for clinicians to observe anatomical structures of the hip joints, even in severely diseased cases.

Evaluation of optimal implant alignment in total hip arthroplasty based on postoperative range of motion simulation

BACKGROUND

Dislocation after total hip arthroplasty is a frequent cause of revision surgery. This study was performed to determine the optimal implant alignment in total hip arthroplasty by simulating the postoperative range of motion.

METHODS

All operations were performed via posterolateral approach using combined anteversion of the stem and cup technique. Maximum range of motion without implant impingement was simulated in 79 replaced hips using postoperative computed tomography and the achievement of the required range of motion defined by previous studies was assessed. Optimal cup and stem alignment for impingement-free range of motion were statistically determined using the receiver operator coefficient curve.

FINDINGS

Cup inclination and anteversion, stem anteversion, and combined anteversion were 37.6°, 20.1°, 26.2°, and 46.3°, respectively. Maximum range of motion in flexion, extension, internal rotation at 90° of flexion, and external rotation were 131.8°, 42.3°, 56.4°, and 64.5°, respectively. Flexion >110°, extension >30°, internal rotation >30° at 90° of flexion, and external rotation >30° were fulfilled by 96%, 86%, 92%, and 96% of all replaced hips, respectively. Optimal implant alignment for impingement-free range of motion was 34°-43° of cup inclination, 18°-26° of cup anteversion, 17°-29° of stem anteversion, and 35°-56° of combined anteversion. Both cup and stem anteversion showed significant relationship with postoperative range of motion.

INTERPRETATION

Surgeons could gain valuable insights into optimal cup and stem alignment to perform postoperative range of motion simulations.

The Accuracy of Three-Dimensional CT Scan Software in Predicting Prosthetic Utilization in Total Shoulder Arthroplasty

INTRODUCTION

Recent innovations in shoulder arthroplasty include three-dimensional (3D) CT software imaging that can be used to predict which prosthetic implants will be used intraoperatively. Correct prediction of the implants may optimize supply chain logistics for the surgeon, hospital, ambulatory surgery center, and the implant company. The purpose of this study was to examine a single surgeon's experience with this software to determine its predictive accuracy in determining which implants would be used intraoperatively.

METHODS

A retrospective review of patients undergoing total shoulder arthroplasty (TSA) performed by a single surgeon was performed. Inclusion criteria were patients undergoing anatomic (aTSA) or reverse (rTSA) TSA examined preoperatively with the 3D CT planning software. A chart review was performed to compare the accuracy of the preoperative plan in predicting the actual prostheses implanted at surgery.

RESULTS

Two hundred seventy-eight shoulders from 260 patients were included. One hundred fifty-one shoulders underwent aTSA, and 127 shoulders underwent rTSA. The surgeon was able to predict the type of arthroplasty (anatomic versus reverse) implanted in 269 of 278 (97%) shoulders. Using the 3D CT software, the surgeon was able to predict all the implants implanted in 68 shoulders (24%). For aTSA, 3D CT imaging successfully predicted all implants implanted in 43 shoulders (28%), glenoid implants implanted in 120 of 148 shoulders (81%), and humeral implants implanted in 54 shoulders (36%). For rTSA, 3D CT imaging successfully predicted all implants implanted in 26 shoulders (20%), glenoid implants implanted in 106 shoulders (83%), and humeral implants implanted in 39 shoulders (31%).

CONCLUSIONS

The 3D CT software combined with surgeon's judgment provided a high accuracy (97%) in determining the type of arthroplasty, a moderately high accuracy in determining the glenoid implants (81% to 83%), a low accuracy in determining humeral implants (31% to 36%), and a low accuracy in determining all prostheses used for each surgery (20% to 28%).

LEVEL OF EVIDENCE

LOE IV-Diagnostic Case Series.

Minimum 10 years clinical results of an anatomical short stem with a proximal hydroxyapatite coating

OBJECTIVES

The CentPillar stem (Stryker Orthopaedics), an anatomical short stem, was designed to match the proximal femoral canal shape in both patients with normal hips and those with developmental dysplasia of the hip (DDH). The long-term outcomes of the CentPillar stem was examined herein.

METHODS

In total, 222 hips that underwent total hip arthroplasty using the CentPillar stem were analyzed. DDH was the main reason for surgery (79.3%). Implant survivorship was assessed using Kaplan-Meier analysis. For radiographic evaluation, stress shielding was assessed using the Engh classification. For functional evaluation, patients' ability to perform deep hip bending activities (sit on the floor, squat, and sit straight) was assessed.

RESULTS

During a median follow-up of 13.1 years, 2 stem revisions were performed (aseptic loosening and late hematogenous periprosthetic infection), and the overall cumulative implant survival rate was 99.0% at 15 years. In the radiographic evaluation, grade >3 stress shielding was found in only one hip. More than 80% of the patients were able to perform each of the three deep hip bending activities.

CONCLUSIONS

Good fixation at the proximal part of the femur was obtained, and the implant survival rate of the CentPillar stem was excellent during a long-term follow-up.

Three-dimensional pre-operative planning of primary hip arthroplasty: a systematic literature review

Three-dimensional (3D) pre-operative planning in total hip arthroplasty (THA) is being recognized as a useful tool in planning elective surgery, and as crucial to define the optimal component size, position and orientation. The aim of this study was to systematically review the existing literature for the use of 3D pre-operative planning in primary THA.A systematic literature search was performed using keywords, through PubMed, Scopus and Google Scholar, to retrieve all publications documenting the use of 3D planning in primary THA. We focussed on (1) the accuracy of implant sizing, restoration of hip biomechanics and component orientation; (2) the benefits and barriers of this tool; and (3) current gaps in literature and clinical practice.Clinical studies have highlighted the accuracy of 3D pre-operative planning in predicting the optimal component size and orientation in primary THAs. Component size planning accuracy ranged between 34-100% and 41-100% for the stem and cup respectively. The absolute, average difference between planned and achieved values of leg length, offset, centre of rotation, stem version, cup version, inclination and abduction were 1 mm, 1 mm, 2 mm, 4°, 7°, 0.5° and 4° respectively.Benefits include 3D representation of the human anatomy for precise sizing and surgical execution. Barriers include increased radiation dose, learning curve and cost. Long-term evidence investigating this technology is limited.Emphasis should be placed on understanding the health economics of an optimized implant inventory as well as long-term clinical outcomes. Cite this article: EFORT Open Rev 2020;5:845-855. DOI: 10.1302/2058-5241.5.200046.

Can 3D surgical planning and patient specific instrumentation reduce hip implant inventory? A prospective study

BACKGROUND

Modern designs of joint replacements require a large inventory of components to be available during surgery. Pre-operative CT imaging aids 3D surgical planning and implant sizing, which should reduce the inventory size and enhance clinical outcome. We aimed to better understand the impact of the use of 3D surgical planning and Patient Specific Instrumentation (PSI) on hip implant inventory.

METHODS

An initial feasibility study of 25 consecutive cases was undertaken to assess the discrepancy between the planned component sizes and those implanted to determine whether it was possible to reduce the inventory for future cases. Following this, we performed a pilot study to investigate the effect of an optimized inventory stock on the surgical outcome: we compared a group of 20 consecutive cases (experimental) with the 25 cases in the feasibility study (control). We assessed: (1) accuracy of the 3D planning system in predicting size (%); (2) inventory size changes (%); (3) intra and post-operative complications.

RESULTS

The feasibility study showed variability within 1 size range, enabling us to safely optimize inventory stock for the pilot study. (1) 3D surgical planning correctly predicted sizes in 93% of the femoral and 89% of the acetabular cup components; (2) there was a 61% reduction in the implant inventory size; (3) we recorded good surgical outcomes with no difference between the 2 groups, and all patients had appropriately sized implants.

CONCLUSIONS

3D planning is accurate in up to 95% of the cases. CT-based planning can reduce inventory size in the hospital setting potentially leading to a reduction in costs.

An increase in the risk of excessive femoral anteversion for relatively younger age and types of femoral morphology in total hip arthroplasty with direct anterior approach

BACKGROUND

Direct anterior approach (DAA) is known to diminish a dislocation risk and widely used for total hip arthroplasty (THA). On the other hand, anterior dislocation due to increasing stem anteversion and cup anteversion is an important complication.

METHODS

A retrospective analysis of 140 (male/female: 10/130) consecutive patients with 149 hips who had undergone primary THA in the period between 2011 and 2015 was conducted. Factors including age, gender, body mass index (BMI), and primary diagnosis were examined. Radiographic parameters including the Dorr proximal femoral types, cortical thickness index (CTI), canal flare index (CFI), and canal to calcar index were measured. To investigate relationships of combined anteversion (CA) with other factors, a stepwise regression analysis was performed.

RESULTS

The mean age, BMI, and the follow-up period were 69.5 ± 10.3 years old, 23.4 ± 3.4 kg/m², and 48.2 ± 13.6 months, respectively. The proximal femurs of the 149 hips were categorized as Dorr type A ( n = 33), type B ( n = 110), and type C ( n = 6). The mean CTI, CFI, and canal to calcar index were 0.56 ± 0.08, 4.15 ± 0.83, and 0.46 ± 0.09, respectively. The mean stem varus angle, radiographic inclination, and radiographic anteversion of the cup were 0.19 ± 1.63, 40.9 ± 6.0°, and 12.7 ± 3.5°, respectively. Stem anteversion measured by computed tomography (CT) axial image was 17.7 ± 12.0°. The mean radiographic inclination, radiographic anteversion (CTRA), and anatomical anteversion analyzed by Kyocera 3-D template were 40.8 ± 7.2°, 19.8 ± 6.6°, and 28.8 ± 10.0°, respectively. Mean CA defined as the sum of stem anteversion and CTRA was 37.5 ± 14.5°. To investigate relationships of CA with other factors, a stepwise regression analysis was performed and resulted in a model with age ( p < 0.001). In the same way, stem anteversion and CTRA were analyzed; only for the stem anteversion, the stepwise selection process resulted in a model with age ( p < 0.001) and the Dorr types ( p = 0.047).

CONCLUSION

The risk of excessive femoral anteversion increases for relatively younger age and for types of femoral morphology according to the Dorr classification, moreover with an increase of CA in DAA-THA with cementless tapered-wedge stem.