Adaptive filtering, modelling and classification of knee joint vibroarthrographic signals for non-invasive diagnosis of articular cartilage pathology

Interpretation of vibrations or sound signals emitted from the patellofemoral joint during movement of the knee, also known as vibroarthrography (VAG), could lead to a safe, objective, and non-invasive clinical tool for early detection, localisation, and quantification of articular cartilage disorders. In this study with a reasonably large database of VAG signals of 90 human knee joints (51 normal and 39 abnormal), a new technique for adaptive segmentation based on the recursive least squares lattice (RLSL) algorithm was developed to segment the non-stationary VAG signals into locally-stationary components; the stationary components were then modelled autoregressively, using the Burg-Lattice method. Logistic classification of the primary VAG signals into normal and abnormal signals (with no restriction on the type of cartilage pathology) using only the AR coefficients as discriminant features provided an accuracy of 68.9% with the leave-one-out method. When the abnormal signals were restricted to chondromalacia patella only, the classification accuracy rate increased to 84.5%. The effects of muscle contraction interference (MCI) on VAG signals were analysed using signals from 53 subjects (32 normal and 21 abnormal), and it was found that adaptive filtering of the MCI from the primary VAG signals did not improve the classification accuracy rate. The results indicate that VAG is a potential diagnostic tool for screening for chondromalacia patella.     

Knee joint vibroarthrography of asymptomatic subjects during loaded flexion-extension movements

Vibroarthrography (VAG) has been proposed as a diagnostic tool for knee joint disorders. The methodological aspects of VAG assessments, such as sensor placement and outcomes measures have not been thoroughly investigated. The aim of this study was to investigate the effects of sensor placement and the effect of the knee extension and flexion movements using different added loads on six parameters extracted from VAGs of asymptomatic subjects. We hypothesized non-uniform distribution of VAGs during knee movement resulting from the underlying biomechanical features. Linear mixed model statistical analysis was carried out. Interpolation maps of the parameters were created to illustrate their spatial distribution over the knee joint. All parameters were significantly related to load, movement type, and location. There were significant load × movement type and location × movement type interaction for ARV and MPF, respectively. Further, the topographical maps of the six VAG parameters differed underlining the complementary information of the six VAG outcomes. The non-uniformity of the spatial distribution, movement type, and the knee joint load dependency of the investigate parameters open for new uses of VAG for joint diagnostic purposes including activities of daily living.

Open image in new window

Graphical abstract Visual summary of the main findings of the study. This figure shows that knee flexion-extension movement with various added external loads (0–1.25–2.5–5 kg) were investigated, vibroarthrogaphic (VAG) signals from eight locations over the knee were recorded, linear and nonlinear analyses were conducted revealing non-uniformity of the spatial distribution of the VAGs.

     

Screening of knee-joint vibroarthrographic signals using statistical parameters and radial basis functions

Externally detected vibroarthrographic (VAG) signals bear diagnostic information related to the roughness, softening, breakdown, or the state of lubrication of the articular cartilage surfaces of the knee joint. Analysis of VAG signals could provide quantitative indices for noninvasive diagnosis of articular cartilage breakdown and staging of osteoarthritis. We propose the use of statistical parameters of VAG signals, including the form factor involving the variance of the signal and its derivatives, skewness, kurtosis, and entropy, to classify VAG signals as normal or abnormal. With a database of 89 VAG signals, screening efficiency of up to 0.82 was achieved, in terms of the area under the receiver operating characteristics curve, using a neural network classifier based on radial basis functions.     

基于PCNN-LSTM神经网络的膝关节摆动信号分类识别

膝关节摆动(VAG)信号是指膝关节屈曲或伸展时发出的声音或振动信号,可灵敏、客观地描述膝关节的健康状态,在膝关节疾病的无创检测中具有重要作用.现有的对VAG信号正常和异常分类方法自动化程度低,且分类准确度较低,总体性能有待进一步提升.因此,提出一种基于改进卷积神经循环网络(PCNN-LSTM)的VAG信号分类算法.首先...     

Analysis of Vibroarthrographic Signals with Features Related to Signal Variability and Radial-Basis Functions

Knee-joint sounds or vibroarthrographic (VAG) signals contain diagnostic information related to the roughness, softening, breakdown, or the state of lubrication of the articular cartilage surfaces. Objective analysis of VAG signals provides features for pattern analysis, classification, and noninvasive diagnosis of knee-joint pathology of various types. We propose parameters related to signal variability for the analysis of VAG signals, including an adaptive turns count and the variance of the mean-squared value computed during extension, flexion, and a full swing cycle of the leg, for the purpose of classification as normal or abnormal, that is, screening. With a database of 89 VAG signals, screening efficiency of up to 0.8570 was achieved, in terms of the area under the receiver operating characteristics curve, using a neural network classifier based on radial-basis functions, with all of the six proposed features. Using techniques for feature selection, the turns counts for the flexion and extension parts of the VAG signals were chosen as the top two features, leading to an improved screening efficiency of 0.9174. The proposed methods could lead to objective criteria for improved selection of patients for clinical procedures and reduce healthcare costs.     

Representation of fluctuation features in pathological knee joint vibroarthrographic signals using kernel density modeling method

This article applies advanced signal processing and computational methods to study the subtle fluctuations in knee joint vibroarthrographic (VAG) signals. Two new features are extracted to characterize the fluctuations of VAG signals. The fractal scaling index parameter is computed using the detrended fluctuation analysis algorithm to describe the fluctuations associated with intrinsic correlations in the VAG signal. The averaged envelope amplitude feature measures the difference between the upper and lower envelopes averaged over an entire VAG signal. Statistical analysis with the Kolmogorov–Smirnov test indicates that both of the fractal scaling index (p = 0.0001) and averaged envelope amplitude (p = 0.0001) features are significantly different between the normal and pathological signal groups. The bivariate Gaussian kernels are utilized for modeling the densities of normal and pathological signals in the two-dimensional feature space. Based on the feature densities estimated, the Bayesian decision rule makes better signal classifications than the least-squares support vector machine, with the overall classification accuracy of 88% and the area of 0.957 under the receiver operating characteristic (ROC) curve. Such VAG signal classification results are better than those reported in the state-of-the-art literature. The fluctuation features of VAG signals developed in the present study can provide useful information on the pathological conditions of degenerative knee joints. Classification results demonstrate the effectiveness of the kernel feature density modeling method for computer-aided VAG signal analysis.     

Computer-aided diagnosis of knee-joint disorders via vibroarthrographic signal analysis: a review

The knee is the lower-extremity joint that supports nearly the entire weight of the human body. It is susceptible to osteoarthritis and other knee-joint disorders caused by degeneration or loss of articular cartilage. The detection of a knee-joint abnormality at an early stage is important, because it helps increase therapeutic options that may slow down the degenerative process. Imaging-based arthrographic modalities can provide anatomical images of the joint cartilage surfaces, but fail to demonstrate the functional integrity of the cartilage. Knee-joint auscultation, by means of recording the vibroarthrographic (VAG) signal during bending motion of a knee, could be used to develop a noninvasive diagnostic tool. Computer-aided analysis of VAG signals could provide quantitative indices for screening of degenerative conditions of the cartilage surface and staging of osteoarthritis. In addition, the diagnosis of knee-joint pathology by means of VAG signal analysis may reduce the number of semi-invasive diagnostic arthroscopic examinations. This article reviews studies related to VAG signal analysis, first summarizing the pilot studies that demonstrated the diagnostic potential of knee-joint auscultation for the detection of degenerative diseases, and then describing the details of recent progress in analysis of VAG signals using temporal analysis, frequency-domain analysis, time-frequency analysis, and statistical modeling. The decision-making methods used in the related studies are summarized, followed by a comparison of the diagnostic performance achieved by different pattern classifiers. The final section is a perspective on the future and further development of VAG signal analysis.     

Characteristics of sensori-motor interaction in the primary and secondary somatosensory cortices in humans: a magnetoencephalography study

We studied sensori-motor interaction in the primary (SI) and secondary somatosensory cortex (SII) using magnetoencephalography. Since SII in both hemispheres was activated following unilateral stimulation, we analyzed SIIc (contralateral to stimulation) as well as SIIi (ipsilateral to stimulation). Four tasks were performed in human subjects in which a voluntary thumb movement of the left or right hand was combined with electrical stimulation applied to the index finger of the left or right hand: L(M)-L(S) (movement of the left thumb triggered stimulation to the left finger), L(M)-R(S) (movement of the left thumb triggered electrical stimulation to the right finger), R(M)-R(S) (movement of the right thumb triggered electrical stimulation to the right finger), and R(M)-L(S) (movement of the right thumb triggered electrical stimulation to the left finger). Stimulation to the index finger only (S condition) was also recorded. In SI, the amplitude of N20m and P35m was significantly attenuated in the R(M)-R(S) and L(M)-L(S) tasks compared with the S condition, but that for other tasks showed no change, corresponding to a conventional gating phenomenon. In SII, the R(M)-L(S) task significantly enhanced the amplitude of SIIc but reduced that of SIIi compared with the S condition. The L(M)-L(S) and R(M)-R(S) tasks caused a significant enhancement only in SIIi. The L(M)-R(S) task enhanced the amplitude only in SIIc. The laterality index showed that SII modulation with voluntary movement was more dominant in the hemisphere ipsilateral to movement but was not affected by the side of stimulation. These results provided the characteristics of activities in somatosensory cortices, a simple inhibition in SI but complicated changes in SII depending on the side of movement and stimulation, which may indicate the higher cognitive processing in SII.     

Automatic de-noising of knee-joint vibration signals using adaptive time-frequency representations

A novel de-noising method for improving the signal-to-noise ratio of knee-joint vibration signals (also known as vibro-arthrographic (VAG) signals) is proposed. The de-noising methods considered are based on signal decomposition techniques, such as wavelets, wavelet packets and the matching pursuit (MP) method. Performance evaluation with synthetic signals simulated with the characteristics expected of VAG signals indicates good de-noising results with the MP method. Statistical pattern classification of non-stationary signal features extracted from time-frequency distributions of 37 (19 normal and 18 abnormal) MP method-de-noised VAG signals shows a sensitivity of 83.3%, a specificity of 84.2% and an overall accuracy of 83.8%.     

Far-field recording of compound muscle action potentials produced during brachial plexus stimulation.

In 15 normal subjects, compound muscle action potentials (CMAPs) produced during electrical stimulation of the brachial plexus were obtained at a distance (far field) from their sources (upper-limb muscles). These far-field CMAPs were recorded at the forehead, neck, and tip of the middle finger and were referred to a relatively inactive region (contralateral forearm or knee). At the proximal recording site (neck or forehead), the far-field CMAPs were generally triphasic in appearance when all of the muscles of the upper limb were activated during brachial plexus stimulation. The initial component was a negative wave; this was followed by a positive component which terminated in a negative potential. At the distal recording site (tip of middle finger), the far-field CMAPs usually had a biphasic appearance, with an initial positivity followed by a period of negativity. The contours of the major components of the far-field CMAPs were frequently interrupted by a series of small amplitude negative and positive peaks or inflections.     

Temporal discrimination of two passive movements in humans: a new psychophysical approach to assessing kinaesthesia

Percutaneous electrical stimulation of the motor point of the first dorsal interosseous muscle (FDI) was used to produce a non-painful contraction of the FDI muscle that caused index finger abduction movement but no radiating cutaneous paraesthesias or sharp sensations localized to joints. Pairs of stimuli separated by different time intervals were given and subjects were asked to report whether they perceived a single or a double index finger abduction movement. The threshold value was the shortest interval for which the subjects reported two separate index finger abduction movements. Temporal discrimination movement thresholds (TDMT) were measured for both right and left hand. To assess the possible role of muscle and cutaneous afferents in temporal discrimination, we investigated the effects of high-frequency (20 Hz) electrical stimulation of the right ulnar and radial nerves on TDMT. In humans, muscle afferents from FDI are supplied by the ulnar nerve whereas the cutaneous territory overlying the muscle and joint is supplied by the radial and median nerves. Threshold values were not significantly different for right (75.1 ms) and left (75.6 ms) hands. During ulnar and to a lesser extent during radial nerve stimulation, TDMT values were significantly increased (119.2 and 93.5 ms, respectively) compared with baseline conditions (78.0 ms) whereas no changes were observed during median nerve stimulation (80.5 ms). These results suggest that muscle, and in part cutaneous, afferents contribute to temporal discrimination of a dual movement. The technique may provide a useful way of measuring temporal discrimination of kinaesthetic inputs in humans.     

Digital envelope detector for blood pressure measurement using an oscillometric method

Blood pressure measurement in the finger artery offers some advantages compared with that in the brachial artery. However, volume oscillometric signals obtained from finger artery measurement are often influenced by motion artefact due to respiration, speaking, involuntary or voluntary movement, etc. In this paper, we developed a digital envelope detector to detect the maximum oscillation criterion in blood pressure measurement for the first time. The digital envelope detector is robust to noise signals generated by motion artefact and filters out the carrier frequency efficiently. To verify the feasibility of our method, we measured blood pressure for eight subjects using our developed system. The results were compared with the auscultation method. In the case of using a digital envelope detector, we could reduce the mean difference error and standard deviation by 30-40%. Our proposed digital envelope detector is a useful tool to improve the accuracy of blood pressure measurement in finger artery.     

Cycle-to-cycle control of swing phase of paraplegic gait induced by surface electrical stimulation

Parameterised swing phase of gait in paraplegics was obtained using surface electrical stimulation of the hip flexors, hamstrings and quadriceps; the hip flexors were stimulated to obtain a desired hip angle range, the hamstrings to provide foot clearance in the forward swing, and the quadriceps to acquire knee extension at the end of the swing phase. We report on two main aspects; optimisation of the initial stimulation parameters, and parameter adaption (control). The initial stimulation patterns were experimentally optimised in two paraplegic subjects using a controlled stand device, resulting in an initial satisfactory swinging motion in both subjects. Intersubject differences appeared in the mechanical output (torque joint) per muscle group. During a prolonged open-loop controlled trial with the optimised but unregulated stimulation onsets and burst duration for the three muscle groups, the hip angle range per cycle initially increased above the desired value and subsequently decreased below it. The mechanical performance of the hamstrings and quadriceps remained relatively unaffected. A cycle-to-cycle controller was then designed, operating on the basis of the hip angle ranges obtained in previous swings. This controller successfully adapted the burst duration of the hip flexors to maintain the desired hip angle range.     

Tonic neuronal activation during simple and complex finger movements analyzed by DC-magnetoencephalography

Functional neuroimaging techniques map neuronal activation indirectly via local concomitant cortical vascular/metabolic changes. In a complementary approach, DC-magnetoencephalography measures neuronal activation dynamics directly, notably in a time range of the slow vascular/metabolic response. Here, using this technique neuronal activation dynamics and patterns for simple and complex finger movements are characterized intraindividually: in 6/6 right-handed subjects contralateral prolonged (30 s each) complex self-paced sequential finger movements revealed stronger field amplitudes over the pericentral sensorimotor cortex than simple movements. A consistent lateralization for contralateral versus ipsilateral finger movements was not found (4/6). A subsequent sensory paradigm focused on somatosensory afferences during the motor tasks and the reliability of the measuring technique. In all six subjects stable sustained neuronal activation during electrical median nerve stimulation was recorded. These neuronal quasi-tonic activation characteristics provide a new non-invasive neurophysiological measure to interpret signals mapped by functional neuroimaging techniques.     

Kinematic and dynamic synergies of human precision-grip movements

We analyzed the adaptability of human thumb and index finger movement kinematics and dynamics to variations of precision grip aperture and movement velocity. Six subjects performed precision grip opening and closing movements under different conditions of movement velocity and movement aperture (thumb and index finger tip-to-tip distance). Angular motion of the thumb and index finger joints was recorded with a CyberGlove and a three-dimensional biomechanical model was used for solving the inverse dynamics problem during precision grip movements, i.e., for calculating joint torques from experimentally obtained angular variations. The time-varying joint angles and joint torques were analyzed by principal-component analysis to quantify the contributions of individual joints in kinematic and dynamic synergies. At the level of movement kinematics, we found subject-specific angular contributions. However, the adaptation to large aperture, achieved by an increase of the relative contribution of the proximal joints, was subject-invariant. At the level of movement dynamics, the adaptation of thumb-index finger movements to task constraints was similar among all subjects and required the linear scaling of joint torques, the synchronization of joint torques under high velocity conditions, and a flexible redistribution of joint torques between the proximal joint of the thumb and that of the index finger. This work represents one of the first attempts at calculating the joint torques during human precision-grip movements and indicates that the dynamic synergies seem to be remarkably simple compared with the synergies found for movement kinematics.     

The movement-induced modulation in discriminability between cutaneous nonpainful stimuli depends on test stimulus intensity

The purpose of this study was to find out whether the finger movement-induced modulation of cutaneous discrimination thresholds varies with the intensity level of the test stimulation in various movement conditions. The effect of active and passive finger movement on cutaneous sensitivity to nonpainful electrical stimulation of threshold and suprathreshold intensity was studied in human subjects. The detection threshold and the just-noticeable amplitude difference (discrimination threshold) at two suprathreshold intensities (3× and 10× detection threshold) were determined using a forced-choice paradigm before and after (controls) or during finger movement. In one condition the stimuli were applied just prior to the movement. The finger was actively or passively moved at the frequency of 1.5 Hz or 3 Hz, and the test stimuli were applied to the moving finger, except in one condition to the contralateral finger. The contralateral condition was used to rule out vigilance- and attention-related mechanisms as a cause of sensitivity changes. Active as well as passive movement of the finger produced a significant increase in the detection threshold during the movement. Also just prior to the movement the detection threshold was increased. Suprathreshold discrimination thresholds in the moving finger were not significantly changed during or just prior to the active movement, whereas during passive movement the discrimination threshold to suprathreshold level stimulation was significantly decreased. When test stimuli were applied to the finger contralateral to the actively moving finger, neither the detection threshold nor the discrimination between stimuli of suprathreshold intensities were significantly changed. The results indicate that both active and passive movement can regionally suppress cutaneous sensitivity to threshold level stimuli, whereas discrimination between cutaneous stimuli of suprathreshold intensities can be improved or be unchanged during passive or active movement, respectively. The suppression of threshold level stimuli may be explained by both the activation of afferent-induced inhibition and corollary efferent barrage from motor to sensory areas of the brain, whereas afferent-induced mechanisms seem to have a predominant role in the improvement of discriminability between suprathreshold stimuli. The suppression of threshold sensitivity and the enhancement (or lack of suppression) of suprathreshold discriminability may be beneficial during movement allowing suppression of noise (e.g., skin receptor activation induced by moving finger) and allowing better discrimination of behaviorally relevant (suprathreshold) stimuli (e.g. discrimination of environmental stimuli during movement).     

Improvement in step clearance via calf muscle stimulation

The aim is to study the influence of electrically stimulated calf muscles on the effectiveness of the swinging leg movement. The study is carried out with a group of patients with incomplete spinal cord injuries both under stationary conditions and during cruth-assisted walking. Before stimulation is applied to the ankle plantar flexors, the knee extensors are inactivated. In each cycle, after ankle plantar flexor stimulation, peroneal stimulation is started, triggering the flexion reflex. From a biomechanical point of view. functional electrical stimulation (FES) of the ankle plantar flexors results in increased ground clearance of the lower extremity. Additionally, the FES-assisted lifting of the heel results in the elimination of extensor tone and thus shortens the swing time.     

Short-latency afferent inhibition during selective finger movement

During individual finger movement, two opposite phenomena occur at the level of the central nervous system that could affect other intrinsic hand muscle representations, unintentional co-activation, and surround inhibition (SI). At rest, excitability in the motor cortex (M1) is inhibited at about 20 ms after electric stimulation of a peripheral nerve [short-latency afferent inhibition (SAI)]. We sought to determine whether SAI changes during selective index finger movement. Effects were measured by the response to transcranial magnetic stimulation in two functionally distinct target muscles of the hand [abductor digiti minimi muscle (ADM), first dorsal interosseus muscle (FDI)]. An increase in SAI in the ADM during index finger movement compared to at rest could help explain the genesis of SI. Electrical stimulation was applied to either the little finger (homotopic for ADM, heterotopic for FDI) or the index finger (heterotopic for ADM, homotopic for FDI). During index finger movement, homotopic SAI was present only in the ADM, and the effect of peripheral stimulation was greater when there was less co-activation. Heterotopic SAI found at rest disappeared with movement. We conclude that during movement, homotopic SAI on the muscle in the surround of the intended movement may contribute to SI.     

Long-latency afferent inhibition during selective finger movement

Stimulation of a peripheral nerve of a hand at rest modulates excitability in the motor cortex and, in particular, leads to inhibition when applied at an interval of approximately 200 ms (long-latency afferent inhibition; LAI). Surround inhibition (SI) is the process that inhibits neighboring muscles not involved in a particular task. The neuronal mechanisms of SI are not known, and it is possible that LAI might contribute to it. Using transcranial magnetic stimulation (TMS) with and without movement of the index finger, the motor-evoked potentials (MEPs) were measured of two functionally distinct target muscles of the hand (abductor digiti minimi muscle = ADM, 1st dorsal interosseus muscle = FDI). Electrical stimulation was applied 180 ms before TMS to either the fifth finger or the index finger. Both homotopic and heterotopic finger stimulation resulted in LAI without movement. With index finger movement, motor output further decreased with homo- and heterotopic stimulation in the ADM. In the moving FDI, however, there was no change with either homo- or heterotopic stimulation. Additionally, in the unstimulated movement trials, LAI increased with the amount of unintentional co-activation that occurred despite attempts to maintain the ADM at rest. However, with finger stimulation added, there were almost no increased MEPs despite co-activation. These findings suggest that LAI increases during movement and can enhance SI.     

Defective temporal discrimination of passive movements in Parkinson's disease

Perception of limb position and motion is abnormal in Parkinson's disease (PD). Despite the fact that the processing of proprioceptive inputs is inherently temporal, most studies have assessed spatial aspects of proprioception in PD patients. Here, we use a recently described method to test whether deficits also exist in temporal discrimination of proprioceptive inputs. We induced index finger abduction or wrist flexion through percutaneous electrical stimulation of the motor point of the first dorsal interosseous muscle (FDI) or the flexor carpii radialis (FCR), respectively. Twelve patients with unilateral bradykinetic-rigid PD and 12 healthy subjects were asked to report whether pairs of stimuli separated by different time intervals produced single or double index finger abduction movement or wrist flexion. The shortest interval at which subjects reported two separated movements was considered as temporal movement discrimination threshold. Results showed that thresholds were significantly higher in PD patients than in control subjects for both FCR and FDI muscle, thus demonstrating for the first time that temporal proprioceptive processing is altered in PD.