A Glove-Based Form Factor for Collecting Joint Acoustic Emissions: Design and Validation

Quantifying Asymmetry between Medial and Lateral Compartment Knee Loading Forces using Acoustic Emissions

OBJECTIVE

Osteoarthritis is the most common type of knee arthritis that can be affected by excessive and compressive loads and can affect one or more compartments of the knee: medial, lateral, and patellofemoral. The medial compartment tends to be the most vulnerable to injuries and research suggests that a better understanding of the medial to lateral load distribution conditions could provide insights to the quantitative usage of knee compartments in activities of daily life.

METHODS

To that end, we present a novel method to quantify the directional bias of asymmetry between the medial and lateral compartment knee joint load by recording knee acoustical emissions and analyzing them using a deep neural network in a subject independent model. We placed four miniature contact microphones on the medial and lateral sides of the patella on both the left and right leg. We compared the handcrafted audio features with the automated features extracted from the convolutional autoencoder which is an unsupervised model that learns the comprehensive representation of the input to determine whether these automated features can better represent the signals characteristic in regard to the structural asymmetry of the knee joint. The input to the convolutional auto encoder (CAE) is a time-frequency representation and different types of these images such as spectrogram and scalogram are compared. We also compared the multi-sensor fusion approach with the performance of a single sensor to determine the robustness of using multiple sensors.

RESULTS

Using a representation learning based approach, we developed a subject independent classification model capable of classifying the asymmetry of the medial and lateral joint load across subjects (accuracy = 83%).

CONCLUSION

The result indicates that wavelet coherence which is the time-frequency correlation of two signals using a wavelet transform yields the best accuracy.

SIGNIFICANCE

These findings suggest that acoustic signals could potentially quantify the direction of medial to lateral load distribution which would broaden the implications for wearable sensing technology for monitoring cartilage health and factors responsible for cartilage breakdown and assessing appropriate rehabilitation exercises without overloading on one side.

Acoustic Emissions From Loaded and Unloaded Knees to Assess Joint Health in Patients With Juvenile Idiopathic Arthritis

OBJECTIVE

We studied and compared joint acoustical emissions (JAEs) in loaded and unloaded knees as digital biomarkers for evaluating knee health status during the course of treatment in patients with juvenile idiopathic arthritis (JIA).

METHODS

JAEs were recorded from 38 participants, performing 10 repetitions of unloaded flexion/extension (FE) and loaded squat exercises. A novel algorithm was developed to detect and exclude rubbing noise and loose microphone artifacts from the signals, and then 72 features were extracted. These features were down-selected based on different criteria to train three logistic regression classifiers. The classifiers were trained with healthy and pre-treatment data and were used to predict the knee health scores of post-treatment data for the same patients with JIA who had a follow-up recording. This knee health score represents the probability of having JIA in a subject (0 for healthy and 1 for arthritis).

RESULTS

Post-treatment knee health scores were lower than pre-treatment scores, agreeing with the clinical records of successful treatment. Regarding loaded versus unloaded knee scores, the squats achieved a higher score on average compared to FEs.

CONCLUSION

In healthy subjects with smooth cartilage, the knee scores of squats and FEs were similar indicating that vibrations from the friction of articulating surfaces do not significantly change by the joint load. However, in subjects with JIA, the scores of squats were higher than the scores of FEs, revealing that these two exercises contain different, possibly clinically relevant, information that could be used to further improve this novel assessment modality in JIA.

Quantifying Signal Quality for Joint Acoustic Emissions Using Graph-Based Spectral Embedding

We present a new method for quantifying signal quality of joint acoustic emissions (JAEs) from the knee during unloaded flexion/extension (F/E) exercises. For ten F/E cycles, JAEs were recorded, in a clinical setting, from 34 healthy knees and 13 with a meniscus tear (n=24 subjects). The recordings were first segmented by F/E cycle and described using time and frequency domain features. Using these features, a symmetric k-nearest neighbor graph was created and described using a spectral embedding. We show how the underlying community structure of JAEs was comparable across joint health levels and was highly affected by artifacts. Each F/E cycle was scored by its distance from a diverse set of manually annotated, clean templates and removed if above the artifact threshold. We validate this methodology by showing an improvement in the distinction between the JAEs of healthy and injured knees. Graph community factor (GCF) was used to detect the number of communities in each recording and describe the heterogeneity of JAEs from each knee. Before artifact removal, there was no significant difference between the healthy and injured groups due to the impact of artifacts on the community construction. Following implementation of artifact removal, we observed improvement in knee health classification. The GCF value for the meniscus tear group was significantly higher than the healthy group (p<0.01). With more JAE recordings being taken in the clinic and at home, this paper addresses the need for a robust artifact removal method which is necessary for an accurate description of joint health.

A Novel Accelerometer Mounting Method for Sensing Performance Improvement in Acoustic Measurements From the Knee

In this study, we propose a new mounting method to improve accelerometer sensing performance in the 50 Hz-10 kHz frequency band for knee sound measurement. The proposed method includes a thin double-sided adhesive tape for mounting and a 3D-printed custom-designed backing prototype. In our mechanical setup with an electrodynamic shaker, the measurements showed a 13 dB increase in the accelerometer's sensing performance in the 1-10 kHz frequency band when it is mounted with the craft tape under 2 N backing force applied through low-friction tape. As a proof-of-concept study, knee sounds of healthy subjects (n = 10) were recorded. When the backing force was applied, we observed statistically significant (p < 0.01) incremental changes in spectral centroid, spectral roll-off frequencies, and high-frequency (1-10 kHz) root-mean-square (RMS) acceleration, while low-frequency (50 Hz-1 kHz) RMS acceleration remained unchanged. The mean spectral centroid and spectral roll-off frequencies increased from 0.8 kHz and 4.15 kHz to 1.35 kHz and 5.9 kHz, respectively. The mean high-frequency acceleration increased from 0.45 mgRMS to 0.9 mgRMS with backing. We showed that the backing force improves the sensing performance of the accelerometer when mounted with the craft tape and the proposed backing prototype. This new method has the potential to be implemented in today's wearable systems to improve the sensing performance of accelerometers in knee sound measurements.

An Interpretable Experimental Data Augmentation Method to Improve Knee Health Classification Using Joint Acoustic Emissions

The characteristics of joint acoustic emissions (JAEs) measured from the knee have been shown to contain information regarding underlying joint health. Researchers have developed methods to process JAE measurements and combined them with machine learning algorithms for knee injury diagnosis. While these methods are based on JAEs measured in controlled settings, we anticipate that JAE measurements could enable accessible and affordable diagnosis of acute knee injuries also in field-deployable settings. However, in such settings, the noise and interference would be greater than in sterile, laboratory environments, which could decrease the performance of existing knee health classification methods using JAEs. To address the need for an objective noise and interference detection method for JAE measurements as a step towards field-deployable settings, we propose a novel experimental data augmentation method to locate and then, remove the corrupted parts of JAEs measured in clinical settings. In the clinic, we recruited 30 participants, and collected data from both knees, totaling 60 knees (36 healthy and 24 injured knees) to be used subsequently for knee health classification. We also recruited 10 healthy participants to collect artifact and joint sounds (JS) click templates, which are audible, short duration and high amplitude JAEs from the knee. Spectral and temporal features were extracted, and clinical data was augmented in five-dimensional subspace by fusing the existing clinical dataset into experimentally collected templates. Then knee scores were calculated by training and testing a linear soft classifier utilizing leave-one-subject-out cross-validation (LOSO-CV). The area under the curve (AUC) was 0.76 for baseline performance without any window removal with a logistic regression classifier (sensitivity = 0.75, specificity = 0.78). We obtained an AUC of 0.86 with the proposed algorithm (sensitivity = 0.80, specificity = 0.89), and on average, 95% of all clinical data was used to achieve this performance. The proposed algorithm improved knee health classification performance by the added information through identification and collection of common artifact sources in JAE measurements. This method when combined with wearable systems could provide clinically relevant supplementary information for both underserved populations and individuals requiring point-of-injury diagnosis in field-deployable settings.

Wearable and Nearable Biosensors and Systems for Healthcare

Biosensors and systems in the form of wearables and "nearables" (i [...].

Estimating Knee Joint Load Using Acoustic Emissions During Ambulation

Quantifying joint load in activities of daily life could lead to improvements in mobility for numerous people; however, current methods for assessing joint load are unsuitable for ubiquitous settings. The aim of this study is to demonstrate that joint acoustic emissions contain information to estimate this internal joint load in a potentially wearable implementation. Eleven healthy, able-bodied individuals performed ambulation tasks under varying speed, incline, and loading conditions while joint acoustic emissions and essential gait measures-electromyography, ground reaction forces, and motion capture trajectories-were collected. The gait measures were synthesized using a neuromuscular model to estimate internal joint contact force which was the target variable for subject-specific machine learning models (XGBoost) trained based on spectral, temporal, cepstral, and amplitude-based features of the joint acoustic emissions. The model using joint acoustic emissions significantly outperformed (p < 0.05) the best estimate without the sounds, the subject-specific average load (MAE = 0.31 ± 0.12 BW), for both seen (MAE = 0.08 ± 0.01 BW) and unseen (MAE = 0.21 ± 0.05 BW) conditions. This demonstrates that joint acoustic emissions contain information that correlates to internal joint contact force and that information is consistent such that unique cases can be estimated.

A Pilot Study to Assess the Reliability of Sensing Joint Acoustic Emissions of the Wrist

Joint acoustic emission (JAE) sensing has recently proven to be a viable technique for non-invasive quantification indicating knee joint health. In this work, we adapt the acoustic emission sensing method to measure the JAEs of the wrist-another joint commonly affected by injury and degenerative disease. JAEs of seven healthy volunteers were recorded during wrist flexion-extension and rotation with sensitive uniaxial accelerometers placed at eight locations around the wrist. The acoustic data were bandpass filtered (150 Hz-20 kHz). The signal-to-noise ratio (SNR) was used to quantify the strength of the JAE signals in each recording. Then, nine audio features were extracted, and the intraclass correlation coefficient (ICC) (model 3,k), coefficients of variability (CVs), and Jensen-Shannon (JS) divergence were calculated to evaluate the interrater repeatability of the signals. We found that SNR ranged from 4.1 to 9.8 dB, intrasession and intersession ICC values ranged from 0.629 to 0.886, CVs ranged from 0.099 to 0.241, and JS divergence ranged from 0.18 to 0.20, demonstrating high JAE repeatability and signal strength at three locations. The volunteer sample size is not large enough to represent JAE analysis of a larger population, but this work will lay a foundation for future work in using wrist JAEs to aid in diagnosis and treatment tracking of musculoskeletal pathologies and injury in wearable systems.

Validity of and recommendations for knee joint acoustic assessments during different movement conditions

Knee joint sounds contain information on joint health, morphology and loading. These acoustic signals may be elicited by further, as yet unknown factors. By assessing potential elicitors and their relative contributions to the acoustic signal, we investigated the validity of vibroarthrographic assessments during different movement conditions with the aim to derive recommendations for their practical usage. Cross-sectional study. Nineteen healthy participants (24.7 ± 2.8 yrs, 7 females) performed five movements: level walking, descending stairs, standing up, sitting down, and forward lunge. Knee joint sounds were recorded by two microphones (medial tibial plateau, patella). Knee joint kinematics and ground reaction forces were recorded synchronously to calculate knee joint moments (Nm/Kg). The mean amplitude (dB) and the median power frequency (Hz) were determined. A repeated measures mixed model investigated the impact of potential predictors (sagittal, frontal, transverse plane and total knee joint moments, knee angular velocity, age, sex, body mass index (BMI) and Tegner Activity Score (TAS)). Most of the amplitudes variance is explained by between-subject differences (tibia: 66.6%; patella: 75.8%), and of the median power frequencies variance by the movement condition (tibia: 97.6%; patella: 98.9%). The final model revealed several predictor variables for both sensors (tibia: sagittal plane, frontal plane, and total knee joint moments, age, and TAS; patella: sagittal plane knee moments, knee angular velocity, TAS). The standardization of the execution of the activities, a between-group matching of variables and the inclusion of co-variates are recommended to increase the validity of vibroarthrographic measurements during different movement conditions of the knee joint.

Reliability of Vibroarthrography to Assess Knee Joint Sounds in Motion

Knee acoustic emissions provide information about joint health and loading in motion. As the reproducibility of knee acoustic emissions by vibroarthrography is yet unknown, we evaluated the intrasession and interday reliability of knee joint sounds. In 19 volunteers (25.6 ± 2.0 years, 11 female), knee joint sounds were recorded by two acoustic sensors (16,000 Hz; medial tibial plateau, patella). All participants performed four sets standing up/sitting down (five repetitions each). For measuring intrasession reliability, we used a washout phase of 30 min between the first three sets, and for interday reliability we used a washout phase of one week between sets 3 and 4. The mean amplitude (dB) and median power frequency (Hz, MPF) were analyzed for each set. Intraclass correlation coefficients (ICCs (2,1)), standard errors of measurement (SEMs), and coefficients of variability (CVs) were calculated. The intrasession ICCs ranged from 0.85 to 0.95 (tibia) and from 0.73 to 0.87 (patella). The corresponding SEMs for the amplitude were ≤1.44 dB (tibia) and ≤2.38 dB (patella); for the MPF, SEMs were ≤13.78 Hz (tibia) and ≤14.47 Hz (patella). The intrasession CVs were ≤0.06 (tibia) and ≤0.07 (patella) (p < 0.05). The interday ICCs ranged from 0.24 to 0.33 (tibia) and from 0 to 0.82 (patella) for both the MPF and amplitude. The interday SEMs were ≤4.39 dB (tibia) and ≤6.85 dB (patella) for the amplitude and ≤35.39 Hz (tibia) and ≤15.64 Hz (patella) for the MPF. The CVs were ≤0.14 (tibia) and ≤0.08 (patella). Knee joint sounds were highly repeatable within a single session but yielded inconsistent results for the interday reliability.