Screening of knee-joint vibroarthrographic signals using the strict 2-surface proximal classifier and genetic algorithm

Diagnostics of Articular Cartilage Damage Based on Generated Acoustic Signals Using ANN—Part II: Patellofemoral Joint

Cartilage loss due to osteoarthritis (OA) in the patellofemoral joint provokes pain, stiffness, and restriction of joint motion, which strongly reduces quality of life. Early diagnosis is essential for prolonging painless joint function. Vibroarthrography (VAG) has been proposed in the literature as a safe, noninvasive, and reproducible tool for cartilage evaluation. Until now, however, there have been no strict protocols for VAG acquisition especially in regard to differences between the patellofemoral and tibiofemoral joints. The purpose of this study was to evaluate the proposed examination and acquisition protocol for the patellofemoral joint, as well as to determine the optimal examination protocol to obtain the best diagnostic results. Thirty-four patients scheduled for knee surgery due to cartilage lesions were enrolled in the study and compared with 33 healthy individuals in the control group. VAG acquisition was performed prior to surgery, and cartilage status was evaluated during the surgery as a reference point. Both closed (CKC) and open (OKC) kinetic chains were assessed during VAG. The selection of the optimal signal measures was performed using a neighborhood component analysis (NCA) algorithm. The classification was performed using multilayer perceptron (MLP) and radial basis function (RBF) neural networks. The classification using artificial neural networks was performed for three variants: I. open kinetic chain, II. closed kinetic chain, and III. open and closed kinetic chain. The highest diagnostic accuracy was obtained for variants I and II for the RBF 9-35-2 and MLP 10-16-2 networks, respectively, achieving a classification accuracy of 98.53, a sensitivity of 0.958, and a specificity of 1. For variant III, a diagnostic accuracy of 97.79 was obtained with a sensitivity and specificity of 0.978 for MLP 8-3-2. This indicates a possible simplification of the examination protocol to single kinetic chain analyses.

Monitoring deterioration of knee osteoarthritis using vibration arthrography in daily activities

BACKGROUND AND OBJECTIVE

Pathological recognition of knee joint using vibration arthrography (VAG) is increasingly becoming prevailed, due to the non-invasive and non-radiative benefits. However, knee joint health monitoring using VAG signals is a difficult problem, since VAG signals are contaminated by strong motion artifacts (MA) caused by knee movements during daily activities, such as squatting. So far few works have investigated this problem. Existing studies mainly focused on clinical diagnosis of knee disorders for 2-class (normal/abnormal) classification using VAG signals, which are less contaminated by MA in the scene when subjects perform knee extension and flexion movements in seated position. The purpose of this study is to propose a framework to monitor knee joint health during daily activities.

METHODS

In this paper, a general framework is designed to monitor knee joint health, which consists of VAG enhancement, feature extraction and fusion, and classification. VAG enhancement aims to remove MA and irrelevant components of knee joint pathologies in raw VAG signals. Distinctive features from enhanced VAG signals are obtained in feature extraction and fusion. Classification can not only distinguish whether the knee joint is normal or abnormal, but also distinguish the grade of deterioration of knee osteoarthritis.

RESULTS

813 VAG signals from VAG-OA dataset, which is currently the largest VAG dataset, have been collected from medical cases in Xijing Hospital of the Fourth Military Medical University during daily activities. Experimental results on VAG-OA dataset showed that the accuracy of 2-class (normal/abnormal) classification was 95.9% with sensitivity 98.1% and specificity 93.3%. For 5-class classification based on deterioration grades of osteoarthritis (OA), we obtained accuracy 74.4%, sensitivity 52.6% and specificity 78.3%.

CONCLUSION

The VAG-OA dataset can be used not only for knee joint health monitoring but also for clinical diagnosis. The designed framework on VAG-OA dataset has high classification accuracy, which is of great value to monitor knee joint health using VAG signals during daily activities. The results also demonstrate that the designed framework significantly outperforms the baselines and several state-of-the-art methods.

Analysis and multiclass classification of pathological knee joints using vibroarthrographic signals

BACKGROUND AND OBJECTIVE

Vibroarthrography (VAG) is a method developed for sensitive and objective assessment of articular function. Although the VAG method is still in development, it shows high accuracy, sensitivity and specificity when comparing results obtained from controls and the non-specific, knee-related disorder group. However, the multiclass classification remains practically unknown. Therefore the aim of this study was to extend the VAG method classification to 5 classes, according to different disorders of the patellofemoral joint.

METHODS

We assessed 121 knees of patients (95 knees with grade I-III chondromalacia patellae, 26 with osteoarthritis) and 66 knees from 33 healthy controls. The vibroarthrographic signals were collected during knee flexion/extension motion using an acceleration sensor. The genetic search algorithm was chosen to select the most relevant features of the VAG signal for classification. Four different algorithms were used for classification of selected features: logistic regression with automatic attribute selection (SimpleLogistic in Weka), multilayer perceptron with sigmoid activation function (MultilayerPerceptron), John Platt's sequential minimal optimization algorithm implementation of support vector classifier (SMO) and random forest tree (RandomForest). The generalization error of classification algorithms was evaluated by stratified 10-fold cross-validation.

RESULTS

We obtained levels of accuracy and AUC metrics over 90%, more than 93% sensitivity and more than 84% specificity for the logistic regression-based method (SimpleLogistic) for a 2-class classification. For the 5-class method, we obtained 69% and 90% accuracy and AUC respectively, and sensitivity and specificity over 91% and 69%.

CONCLUSIONS

The results of this study confirm the high usefulness of quantitative analysis of VAG signals based on classification techniques into normal and pathological knees and as a promising tool in classifying signals of various knee joint disorders and their stages.

CLASSIFICATION OF NORMAL AND KNEE JOINT DISORDER VIBROARTHROGRAPHIC SIGNALS USING MULTIFRACTALS AND SUPPORT VECTOR MACHINES

The development of reliable Computer Aided Diagnosis (CAD) systems would help in the early detection of Knee Joint Disorder (KJD). In this work, normal and KJD vibroarthrographic (VAG) signals are classified using multifractals and Support Vector Machines (SVM). Multifractal dimension [Formula: see text] is calculated from the VAG signals for various [Formula: see text]-values ([Formula: see text]). Geometrical features are calculated from the multifractal spectrum. The dimension of the feature set is reduced using Principal Component Analysis (PCA). The significant features obtained from the multifractal spectrum are fed as the input to the SVM classifier. The accuracy of the classifier is analyzed using kernels such as linear, quadratic, polynomial and Radial Basis Functions (RBF). The results suggest that VAG signals exhibits the multifractal property. The fluctuations in the normal and abnormal signals are well predicted in small scales of segments of time series. The features such as [Formula: see text] and Mean[Formula: see text] are high in abnormal VAG signals. These features give statistically significant values in differentiating the normal and abnormal subjects ([Formula: see text]). The area under the Receiver Operating Characteristic (ROC) curve is high for polynomial function (0.98). The SVM classifier with polynomial function gives 92.13% of accuracy in differentiating the normal and abnormal subjects. The calculation of multifractal spectrum and geometrical features from VAG signals requires optimization of few parameters, easy to compute, computationally inexpensive, and less time consuming. Hence, the CAD system seems to be clinically significant for the classification of normal and KJD subjects.

Quantification of knee vibroarthrographic signal irregularity associated with patellofemoral joint cartilage pathology based on entropy and envelope amplitude measures

BACKGROUND AND OBJECTIVE

Injury of knee joint cartilage may result in pathological vibrations between the articular surfaces during extension and flexion motions. The aim of this paper is to analyze and quantify vibroarthrographic (VAG) signal irregularity associated with articular cartilage degeneration and injury in the patellofemoral joint.

METHODS

The symbolic entropy (SyEn), approximate entropy (ApEn), fuzzy entropy (FuzzyEn), and the mean, standard deviation, and root-mean-squared (RMS) values of the envelope amplitude, were utilized to quantify the signal fluctuations associated with articular cartilage pathology of the patellofemoral joint. The quadratic discriminant analysis (QDA), generalized logistic regression analysis (GLRA), and support vector machine (SVM) methods were used to perform signal pattern classifications.

RESULTS

The experimental results showed that the patients with cartilage pathology (CP) possess larger SyEn and ApEn, but smaller FuzzyEn, over the statistical significance level of the Wilcoxon rank-sum test (p<0.01), than the healthy subjects (HS). The mean, standard deviation, and RMS values computed from the amplitude difference between the upper and lower signal envelopes are also consistently and significantly larger (p<0.01) for the group of CP patients than for the HS group. The SVM based on the entropy and envelope amplitude features can provide superior classification performance as compared with QDA and GLRA, with an overall accuracy of 0.8356, sensitivity of 0.9444, specificity of 0.8, Matthews correlation coefficient of 0.6599, and an area of 0.9212 under the receiver operating characteristic curve.

CONCLUSIONS

The SyEn, ApEn, and FuzzyEn features can provide useful information about pathological VAG signal irregularity based on different entropy metrics. The statistical parameters of signal envelope amplitude can be used to characterize the temporal fluctuations related to the cartilage pathology.

Joint Motion Quality in Chondromalacia Progression Assessed by Vibroacoustic Signal Analysis

BACKGROUND

Because of the specific biomechanical environment of the patellofemoral joint, chondral disorders, including chondromalacia, often are observed in this articulation. Chondromalacia via pathologic changes in cartilage may lead to qualitative impairment of knee joint motion.

OBJECTIVE

To determine the patellofemoral joint motion quality in particular chondromalacia stages and to compare with controls.

DESIGN

Retrospective, comparative study.

SETTING

Voivodship hospitals, university biomechanical laboratory.

PATIENTS

A total of 89 knees with chondromalacia (25 with stage I; 30 with stage II and 34 with stage III) from 50 patients and 64 control healthy knees (from 32 individuals).

METHODS

Vibroacoustic signal pattern analysis of joint motion quality.

MAIN OUTCOME MEASUREMENTS

For all knees vibroacoustic signals were recorded. Each obtained signal was described by variation of mean square, mean range (R4), and power spectral density for frequency of 50-250 Hz (P1) and 250-450 Hz (P2) parameters.

RESULTS

Differences between healthy controls and all chondromalacic knees as well as chondromalacia patellae groups were observed as an increase of analyzed parameters (P < .001) with only one exception. No statistically significant difference between control group and stage I of chondromalacia patellae was found. All chondromalacia groups were differentiated by the use of all analyzed parameters (P < .01), whose values correspond to the progress of chondromalacia.

CONCLUSIONS

Chondromalacia generates abnormal vibroacoustic signals, and there seems to be a relationship between the level of signal amplitude as well as frequency and cartilage destruction from the superficial layer to the subchondral bone.

LEVEL OF EVIDENCE

IV.

Age-related impairment of quality of joint motion in vibroarthrographic signal analysis

Aging is associated with degenerative changes in articular surfaces leading to quantitative and qualitative impairment of joint motion. Therefore, the aim of this study is to evaluate an age-related quality of the patellofemoral joint (PFJ) motion in the vibroarthrographic (VAG) signal analysis. Two hundred and twenty individuals were enrolled in this study and divided into five groups according to age. The VAG signals were collected during flexion/extension knee motion using an acceleration sensor and described using four parameters (VMS, P1, P2, and H). We observed that values of parameters VMS, P1, and P2 increase in accordance with the age, but H level decreases. The most significant differences were achieved between the youngest and the oldest participants' groups. Moreover, we show that parameters VMS, P1, and P2 positively correlate with age, contrary to negatively associated H parameter. Our results suggest that the impairment of joint motion is a result of age-related osteoarticular degenerative changes.

Coefficient of Friction Patterns Can Identify Damage in Native and Engineered Cartilage Subjected to Frictional-Shear Stress

The mechanical loading environment encountered by articular cartilage in situ makes frictional-shear testing an invaluable technique for assessing engineered cartilage. Despite the important information that is gained from this testing, it remains under-utilized, especially for determining damage behavior. Currently, extensive visual inspection is required to assess damage; this is cumbersome and subjective. Tools to simplify, automate, and remove subjectivity from the analysis may increase the accessibility and usefulness of frictional-shear testing as an evaluation method. The objective of this study was to determine if the friction signal could be used to detect damage that occurred during the testing. This study proceeded in two phases: first, a simplified model of biphasic lubrication that does not require knowledge of interstitial fluid pressure was developed. In the second phase, frictional-shear tests were performed on 74 cartilage samples, and the simplified model was used to extract characteristic features from the friction signals. Using support vector machine classifiers, the extracted features were able to detect damage with a median accuracy of approximately 90%. The accuracy remained high even in samples with minimal damage. In conclusion, the friction signal acquired during frictional-shear testing can be used to detect resultant damage to a high level of accuracy.

Joint motion quality in vibroacoustic signal analysis for patients with patellofemoral joint disorders

BACKGROUND

Chondromalacia, lateral patellar compression syndrome and osteoarthritis are common patellofemoral joint disorders leading to functional and/or structural disturbances in articular surfaces. The objective of the study was to evaluate their impact on joint motion quality via the vibroacoustic signal generated during joint movement analysis.

METHODS

Seventy-three patients (30 with chondromalacia, 21 with lateral patellar compression syndrome, and 22 with osteoarthritis) and 32 healthy controls were tested during flexion/extension knee motion for vibroacoustic signals using an acceleration sensor. Estimated parameters: variation of mean square (VMS), difference between mean of four maximum and mean of four minimum values (R4), power spectral density for frequency of 50-250 Hz (P1) and 250-450 Hz (P2) were analyzed.

RESULTS

Vibroacoustic signals recorded for particular disorders were characterized by significantly higher values of parameters in comparison to the control group. Moreover, differences were found among the various types of patellofemoral joint disturbances. Chondromalacia and osteoarthritis groups showed differences in all parameters examined. In addition, osteoarthritis patients exhibited differences in VMS, P1 and P2 values in comparison to lateral patellar compression syndrome patients. However, only the value of R4 was found to differ between knees with lateral patellar compression syndrome and those with chondromalacia.

CONCLUSION

Our results suggest that particular disorders are characterized by specific vibroacoustic patterns of waveforms as well as values of analyzed parameters.

Knee joint vibration signal analysis with matching pursuit decomposition and dynamic weighted classifier fusion

Analysis of knee joint vibration (VAG) signals can provide quantitative indices for detection of knee joint pathology at an early stage. In addition to the statistical features developed in the related previous studies, we extracted two separable features, that is, the number of atoms derived from the wavelet matching pursuit decomposition and the number of significant signal turns detected with the fixed threshold in the time domain. To perform a better classification over the data set of 89 VAG signals, we applied a novel classifier fusion system based on the dynamic weighted fusion (DWF) method to ameliorate the classification performance. For comparison, a single leastsquares support vector machine (LS-SVM) and the Bagging ensemble were used for the classification task as well. The results in terms of overall accuracy in percentage and area under the receiver operating characteristic curve obtained with the DWF-based classifier fusion method reached 88.76% and 0.9515, respectively, which demonstrated the effectiveness and superiority of the DWF method with two distinct features for the VAG signal analysis.