Knee Osteoarthritis SCAENet: Adaptive Knee Osteoarthritis Severity Assessment Using Spatial Separable Convolution with Attention-Based Ensemble Networks with Hybrid Optimization Strategy

Osteoarthritis (OA) of the knee is a chronic state that significantly lowers the quality of life for its patients. Early detection and lifetime monitoring of the progression of OA are necessary for preventive therapy. In the course of therapy, the Kellgren and Lawrence (KL) assessment model categorizes the rigidity of OA. Deep techniques have recently been used to increase the precision and effectiveness of OA severity assessments. The training process is compromised by low-confidence samples, which are less accurate than normal ones. In this work, a deep learning-based knee osteoarthritis severity assessment model is recommended to accurately identify the condition in patients. The phases of the designed model are data collection, feature extraction, and prediction. At first, the images are generally gathered from online resources. The gathered images are given into the feature extraction phase. A new model is implemented to predict knee osteoarthritis named Spatial Separable Convolution with Attention-based Ensemble Networks (SCAENet), which includes feature extraction, stacked target-based feature pool generation, and knee osteoarthritis prediction. The feature extraction is done using ResNet, Visual Geometry Group (VGG16), and DenseNet. The stacked target-based feature pool is obtained from the SCAENet. Hence, the stacked target-based feature pool is obtained by the Hybridization of Equilibrium Slime Mould with Bald Eagle Search Optimization (HESM-BESO). Here, the knee osteoarthritis's severity prediction is performed using the dimensional convolutional neural network (1DCNN) technique. The designed SCAENet model is validated with other conventional methods to show high performance.

Characterization of immune microenvironment infiltration and m6A regulator-mediated RNA methylation modification patterns in osteoarthritis

Background

Few studies have been reported the potential role of N6-methyladenosine (m6A) modification in osteoarthritis (OA). We investigated the patterns of m6A modification in the immune microenvironment of OA.

Methods

We evaluated the m6A modification patterns based on 22 m6A regulators in 139 OA samples and systematically associated these modification patterns with immune cell infiltration characteristics. The function of m6A phenotype-related differentially expressed genes (DEGs) was investigated using gene enrichment analysis. An m6A score model was constructed using principal component analysis (PCA), and an OA prediction model was established based on the key m6A regulators. We used real-time PCR analysis to detect the changes of gene expression in the cell model of OA.

Results

Healthy and OA samples showed significant differences in the expression of m6A regulators. Nine key m6A regulators, two m6A modification patterns, m6A-related genes and two gene clusters were identified. Some m6A regulators had a strong correlation with each other. Gene clusters and m6A clusters have high similarity, and cluster A corresponds to a high m6A score. Immunocytes infiltration differed significantly between the two clusters, with the m6A cluster B and gene cluster B having more types of infiltrating immunocytes than cluster A. The predictive model can also predict the progression of OA through m6A regulators expression. The results of real-time PCR analysis showed that the gene expression in the cell model of OA is similar to that of the m6A cluster B.

Conclusions

Our study reveals for the first time the potential regulatory mechanism of m6A modification in the immune microenvironment of OA. This study also sheds new light on the pathogenesis of OA.

Patient and Implant Performance of Satisfied and Dissatisfied Total Knee Arthroplasty Patients

BACKGROUND

Implant migration and altered kinematics have been thought to impact patient-reported outcome measures (PROMs) and postoperative patient satisfaction. In this study comparing satisfied and dissatisfied total knee arthroplasty (TKA) patients, we hypothesized that dissatisfied patients will have greater continuous implant migration and that there will be differences in joint kinematics, objective functional measurements, and PROMs between satisfied and dissatisfied patients.

METHODS

The Knee Society Score Satisfaction Subsection questions regarding satisfaction with function were used at least 6 months postoperation to split 50 patients into satisfied and dissatisfied groups. Patients underwent radiostereometric analysis to evaluate migration and kinematics. A wearable sensor system obtained objective measurements of patient function during timed up and go tests. PROMs were recorded preoperation and postoperation.

RESULTS

No statistically significant differences were found in migration between satisfied and dissatisfied groups. Statistical kinematic differences existed in lateral anteroposterior contact location at 20° and 40° of flexion at 1 year, where the dissatisfied group had more anteriorly located lateral contact. No statistically significant differences were present in objective functional measurements. Satisfied and dissatisfied groups had differing PROMs at 4 timepoints or greater for each questionnaire.

CONCLUSIONS

No differences were found in tibial component migration or objectively measured function between satisfied and dissatisfied patients. Functionally dissatisfied patients had more anteriorly positioned contact on the lateral condyle in early flexion and reported more pain and unmet expectations. These findings suggest that improving the functional satisfaction of TKA requires restoration of kinematics in early flexion and management of patient's pain and expectations.

Monitoring daily shoulder activity before and after reverse total shoulder arthroplasty using inertial measurement units

BACKGROUND

The purpose of this study was to use at-home, portable, continuous monitoring technologies to record arm motion and activity preoperatively and postoperatively after reverse total shoulder arthroplasty (RTSA).

METHODS

Thirty-three patients indicated for RTSA were monitored preoperatively and 3 and 12 months postoperatively. Inertial measurement units were placed on the sternum and upper arm of the operative limb, recording humeral motion relative to the torso for the duration of a waking day. Elevation events per hour (EE/h) > 90°, time spent at >90°, and activity intensity were calculated and compared between time points. Patient-reported outcome measures were also collected at all time points.

RESULTS

At 3 (P = .040) and 12 (P = .010) months after RTSA, patients demonstrated a significantly greater number of EE/h > 90° compared with preoperatively. There were no significant differences (P ≥ .242) in the amount of time spent at different elevation angles at any time point or in arm activity intensity. Overall, 95% of the day was spent at elevation angles < 60°, and 90% of the day was spent in a low- or moderate-intensity state. Pearson correlations demonstrated relationships between forward elevation and the number of EE/h (r = 0.395, P = .001) and the number of EE/h > 90° (r = 0.493, P < .001).

CONCLUSION

After RTSA, patients significantly increase the frequency of arm elevation to higher angles. However, we found no differences in the amount of time spent at different elevation angles. Overall, after RTSA, >95% of the day was spent at elevation angles < 60° and <1% of the day was spent at >90° of elevation.

Sleeve for Knee Angle Monitoring: An IMU-POF Sensor Fusion System

The knee flexion-extension angle is an important variable to be monitored in various clinical scenarios, for example, during physical rehabilitation assessment. The purpose of this work is to develop and validate a sensor fusion system based on a knee sleeve for monitoring of physical therapy. The system consists of merging data from two inertial measurement units (IMUs) and an intensity-variation based Polymer Optical Fiber (POF) curvature sensor using a quaternion-based Multiplicative Extended Kalman Filter (MEKF). The proposed data fusion method is magnetometer-free and deals with sensors' uncertainties through reliability intervals defined during gait. Walking trials were performed by twelve healthy participants using our knee sleeve system and results were validated against a gold standard motion capture system. Additionally, a comparison with other three knee angle estimation methods, which are exclusively based on IMUs, was carried out. The proposed system presented better performance (mean RMSE 3.3 ^°, LFM coefficients, a₁ = 0.99 ± 0.04, a₀ = 0.70 ± 2.29, R² = 0.98 ± 0.01 and ρ_C 0.99) when compared to the other evaluated methods. Experimental results demonstrate the usability and feasibility of our system to estimate knee motion with high accuracy, repeatability, and reproducibility. This wearable system may be suitable for motion assessment in rehabilitation labs in future studies.

Machine Learning Predicts the Fall Risk of Total Hip Arthroplasty Patients Based on Wearable Sensor Instrumented Performance Tests

BACKGROUND

The prevalence of falls affects the wellbeing of aging adults and places an economic burden on the healthcare system. Integration of wearable sensors into existing fall risk assessment tools enables objective data collection that describes the functional ability of patients. In this study, supervised machine learning was applied to sensor-derived metrics to predict the fall risk of patients following total hip arthroplasty.

METHODS

At preoperative, 2-week, and 6-week postoperative appointments, patients (n = 72) were instrumented with sensors while they performed the timed-up-and-go walking test. Preoperative and 2-week postoperative data were used to form the feature sets and 6-week total times were used as labels. Support vector machine and linear discriminant analysis classifier models were developed and tested on various combinations of feature sets and feature reduction schemes. Using a 10-fold leave-some-subjects-out testing scheme, the accuracy, sensitivity, specificity, and area under the receiver-operator curve (AUC) were evaluated for all models.

RESULTS

A high performance model (accuracy = 0.87, sensitivity = 0.97, specificity = 0.46, AUC = 0.82) was obtained with a support vector machine classifier using sensor-derived metrics from only the preoperative appointment. An overall improved performance (accuracy = 0.90, sensitivity = 0.93, specificity = 0.59, AUC = 0.88) was achieved with a linear discriminant analysis classifier when 2-week postoperative data were added to the preoperative data.

CONCLUSION

The high accuracy of the fall risk prediction models is valuable for patients, clinicians, and the healthcare system. High-risk patients can implement preventative measures and low-risk patients can be directed to enhanced recovery care programs.

A four-bar knee joint measurement walking system for prosthesis design

BACKGROUND

Gait analysis is important for the lower limb prosthesis design. Simulating the natural motion of the human knee in different terrains is useful for the design and performance assessment of the prosthetic knee.

OBJECTIVE

This study aimed to propose a four-bar knee joint measurement system which can simulate the natural knee motions to collect the kinetic parameters precisely and analyze the walking characteristics under different terrain conditions.

METHODS

A low-cost four-bar knee joint mechanism was proposed and gait characteristics were assessed on level ground, ascending and descending stairs, and ascending and descending ramp.

RESULTS

The initial knee flexion angle during stair ascent at heel strike is obviously larger than in other walking scenes. The stance phase accounts for 53% of a single gait cycle during stair descent, which is slightly lower than other walking scenarios. The period that both the hindfoot and forefoot contact the ground in ramp descent accounts for 18%, which is less than for the others. While the forefoot contacts the ground in ramp ascent, the maximum vertical ground reaction force of the forefoot occurs when the hindfoot and forefoot simultaneously contact the ground, whereas in other scenarios the forefoot contacts the ground solely.

CONCLUSIONS

The four-bar knee joint can simulate the natural motion of the human knee accurately. The gait characteristics analysis of different walking scenarios indicated that the low-cost four-bar knee joint exoskeleton was suitable for human knee joint simulation.

PyTHang: an open-source wearable sensor system for real-time monitoring of head-torso angle for ambulatory applications

This article presents the realization of a low-cost wearable sensor system and its Python-based software that can measure and record relative head-torso angle, especially in sagittal plane. The system is mainly developed to track head-torso angle during walk in a clinical study. The open-hardware part of the system is composed of a pair of triaxial digital accelerometers, a microprocessor, a Bluetooth module and a rechargeable battery unit. The reception of the transmitted acceleration data, visualization, interactive sensor alignment, angle estimation and data-logging are realized by the developed open-source graphical user interface. The system is tested on a tripod for verification and on a subject for practical demonstration. Developed system can be constructed and used for ambulatory monitoring and analysis of relative head-torso angle. Open-source user interface can be downloaded and developed for further (different) algorithms and device hardware.

Machine learning and wearable sensors at preoperative assessments: Functional recovery prediction to set realistic expectations for knee replacements

Unmet expectations contribute to a high patient dissatisfaction rate following total knee replacement but clinicians currently do not have the tools to confidently adjust expectations. In this study, supervised machine learning was applied to multi-variate wearable sensor data from preoperative timed-up-and-go tests. Participants (n=82) were instrumented three months after surgery and patients showing relevant improvement were designated as "responders" while the remainder were labelled "maintainers". Support vector machine, naïve Bayes, and random forest binary classifiers were developed to distinguish patients using sensor-derived features. Accuracy, sensitivity, specificity, and area under the receiver-operator curve (AUC) were compared between models using ten-fold out-of-sample testing. A high performance using only sensor-derived functional metrics was obtained with a random forest model (accuracy = 0.76 ± 0.11, sensitivity = 0.87 ± 0.08, specificity = 0.57 ± 0.26, AUC = 0.80 ± 0.14) but highly sensitive models were observed using naïve Bayes and SVM models after including patient age, sex, and BMI into the feature set (accuracy = 0.72, 0.73 ± 0.09, 0.12; sensitivity = 0.94, 0.95 ± 0.11, 0.11; specificity = 0.35, 0.37 ± 0.20, 0.18; AUC = 0.80, 0.74 ± 0.07, 0.11; respectfully). Including select patient-reported subjective measures increased the top random forest performance slightly (accuracy = 0.80 ± 0.10, sensitivity = 0.91 ± 0.14, specificity = 0.62 ± 0.23, AUC = 0.86 ± 0.09). The current work has demonstrated that prediction models developed from preoperative sensor-derived functional metrics can reliably predict expected functional recovery following surgery and this can be used by clinicians to help set realistic patient expectations.

The effect of femoral stem collar on implant migration and clinical outcomes following direct anterior approach total hip arthroplasty

AIMS

The direct anterior (DA) approach has been associated with rapid patient recovery after total hip arthroplasty (THA) but may be associated with more frequent femoral complications including implant loosening. The objective of this study was to determine whether the addition of a collar to the femoral stem affects implant migration, patient activity, and patient function following primary THA using the DA approach.

METHODS

Patients were randomized to either a collared (n = 23) or collarless (n = 26) cementless femoral stem implanted using the DA approach. Canal fill ratio (CFR) was measured on the first postoperative radiographs. Patients underwent a supine radiostereometric analysis (RSA) exam postoperatively on the day of surgery and at two, four, six, 12, 26, and 52 weeks postoperatively. Patient-reported outcome measures (Western Ontario and McMaster Universities Osteoarthritis (WOMAC) Index, the 12-item Short Form Health Survey Mental and Physical Score, and University of California, Los Angeles (UCLA) Activity Score) were measured preoperatively and at each post-surgery clinic visit. Activity and function were also measured as the weekly average step count recorded by an activity tracker, and an instrumented timed up-and-go (TUG) test in clinic, respectively.

RESULTS

Comparing the RSA between the day of surgery baseline exam to two weeks postoperatively, subsidence was significantly lower (mean difference 2.23 mm (SD 0.71), p = 0.023) with collared stems, though these patients had a greater CFR (p = 0.048). There was no difference (p = 0.426) in subsidence between stems from a two-week baseline through to one year postoperatively. There were no clinically relevant differences in PROMs; and there was no difference in the change in activity (p = 0.078) or the change in functional capacity (p = 0.664) between the collared stem group and the collarless stem group at any timepoint.

CONCLUSION

Presence of a collar on the femoral stem resulted in reduced subsidence during the first two postoperative weeks following primary THA using the DA approach. However, the clinical implications are unclear, and larger studies examining patient activity and outcomes are required. Cite this article: Bone Joint J 2020;102-B(12):1654-1661.

Machine Learning Groups Patients by Early Functional Improvement Likelihood Based on Wearable Sensor Instrumented Preoperative Timed-Up-and-Go Tests

BACKGROUND

Wearable sensors permit efficient data collection and unobtrusive systems can be used for instrumenting knee patients for objective assessment. Machine learning can be leveraged to parse the abundant information these systems provide and segment patients into relevant groups without specifying group membership criteria. The objective of this study is to examine functional parameters influencing favorable recovery outcomes by separating patients into functional groups and tracking them through clinical follow-ups.

METHODS

Patients undergoing primary unilateral total knee arthroplasty (n = 68) completed instrumented timed-up-and-go tests preoperatively and at their 2-, 6-, and 12-week follow-up appointments. A custom wearable system extracted 55 metrics for analysis and a K-means algorithm separated patients into functionally distinguished groups based on the derived features. These groups were analyzed to determine which metrics differentiated most and how each cluster improved during early recovery.

RESULTS

Patients separated into 2 clusters (n = 46 and n = 22) with significantly different test completion times (12.6 s vs 21.6 s, P < .001). Tracking the recovery of both groups to their 12-week follow-ups revealed 64% of one group improved their function while 63% of the other maintained preoperative function. The higher improvement group shortened their test times by 4.94 s, (P = .005) showing faster recovery while the other group did not improve above a minimally important clinical difference (0.87 s, P = .07). Features with the largest effect size between groups were distinguished as important functional parameters.

CONCLUSION

This work supports using wearable sensors to instrument functional tests during clinical visits and using machine learning to parse complex patterns to reveal clinically relevant parameters.