Biomedical signal processing (in four parts). Part 1. Time-domain methods

Is There Any Effect of Symmetry on Velocity of the Four Swimming Strokes?

The different characteristics of the four swimming strokes affect the interplay between the four limbs, acting as a constraint to the force produced by each hand and foot. The purpose of this study was to analyze the symmetry of force production with a varying number of limbs in action and see its effect on velocity. Fifteen male swimmers performed four all-out bouts of 25-m swims in the four strokes in full-body stroke and segmental actions. A differential pressure system was used to measure the hands/feet propulsive force and a mechanical velocity meter was used to measure swimming velocity. Symmetry index was calculated based on the force values. All strokes and conditions presented contralateral limb asymmetries (ranging from 6.73% to 28% for the peak force and from 9.3% to 35.7% for the mean force). Backstroke was the most asymmetric stroke, followed-up by butterfly, front crawl, and breaststroke. Kicking conditions elicited the higher asymmetries compared with arm-pull conditions. No significant associations were found between asymmetries and velocity. The absence of such association suggests that, to a certain and unknown extent, swimming may benefit from contralateral limb asymmetry.

Reconfigurable hardware implementation of coherent averaging technique for ultrasonic NDT instruments

The paper proposed a novel hardware (FPGA) implementation of the coherent averaging architecture for the reconfigurable ultrasonic NDT system. The proposed hardware architecture uses the addressing based shifting technique for the addition operation and Radix-2 non-restoring algorithm for the division operation. Since the amount of hardware required by the proposed averaging scheme is independent of the number of averages, it supports on-the-fly control on the number of averages. Compared to conventional architecture, it provides 96% reduction in memory storage, 98% reduction in the number of adders, and 32% reduction in the processing time for the case of 64 coherent averages. For the experimentation, the ultrasonic imaging system designed and developed by the authors has been utilized. The developed system further supports dynamic on-line reconfiguration of the analog front-end hardware, real-time data acquisition, real-time hardware based data processing, and data transfer operations. The performance of implemented coherent averaging has been presented by various applications such as removal of RF random false-echoes, smoothing of A-scan waveforms and speckle removal of B-scan images.

Common vs. Distinct Visuomotor Control Deficits in Autism Spectrum Disorder and Schizophrenia

Autism spectrum disorder (ASD) and schizophrenia (SCZ) are neurodevelopmental disorders with partly overlapping clinical phenotypes including sensorimotor impairments. However, direct comparative studies on sensorimotor control across these two disorders are lacking. We set out to compare visuomotor upper limb impairment, quantitatively, in ASD and SCZ. Patients with ASD (N = 24) were compared to previously published data from healthy control participants (N = 24) and patients with SCZ (N = 24). All participants performed a visuomotor grip force-tracking task in single and dual-task conditions. The dual-task (high cognitive load) presented either visual distractors or required mental addition during grip force-tracking. Motor inhibition was measured by duration of force release and from principal component analysis (PCA) of the participant's force-trajectory. Common impairments in patients with ASD and SCZ included increased force-tracking error in single-task condition compared to controls, a further increase in error in dual-task conditions, and prolonged duration of force release. These three sensorimotor impairments were found in both patient groups. In contrast, distinct impairments in patients with ASD included greater error under high cognitive load and delayed onset of force release compared to SCZ. The PCA inhibition component was higher in ASD than SCZ and controls, correlated to duration of force release, and explained group differences in tracking error. In conclusion, sensorimotor impairments related to motor inhibition are common to ASD and SCZ, but more severe in ASD, consistent with enhanced neurodevelopmental load in ASD. Furthermore, impaired motor anticipation may represent a further specific impairment in ASD. Autism Res 2020, 13: 885-896. © 2020 International Society for Autism Research, Wiley Periodicals, Inc. LAY SUMMARY: Autism spectrum disorder (ASD) and schizophrenia (SCZ) are neurodevelopmental disorders with partly overlapping and partly distinct clinical symptoms. Sensorimotor impairments rank among these symptoms, but it is less clear whether they are shared or distinct. In this study, we showed using a grip force task that sensorimotor impairments related to motor inhibition are common to ASD and SCZ, but more severe in ASD. Impaired motor anticipation may represent a further specific impairment in ASD.

Compressive sensing meets time-frequency: An overview of recent advances in time-frequency processing of sparse signals

Compressive sensing is a framework for acquiring sparse signals at sub-Nyquist rates. Once compressively acquired, many signals need to be processed using advanced techniques such as time-frequency representations. Hence, we overview recent advances dealing with time-frequency processing of sparse signals acquired using compressive sensing approaches. The paper is geared towards signal processing practitioners and we emphasize practical aspects of these algorithms. First, we briefly review the idea of compressive sensing. Second, we review two major approaches for compressive sensing in the time-frequency domain. Thirdly, compressive sensing based time-frequency representations are reviewed followed by descriptions of compressive sensing approaches based on the polynomial Fourier transform and the short-time Fourier transform. Lastly, we provide brief conclusions along with several future directions for this field.

Freezing of Gait Detection in Parkinson's Disease: A Subject-Independent Detector Using Anomaly Scores

Freezing of gait (FoG) is common in Parkinsonian gait and strongly relates to falls. Current clinical FoG assessments are patients' self-report diaries and experts' manual video analysis. Both are subjective and yield moderate reliability. Existing detection algorithms have been predominantly designed in subject-dependent settings. In this paper, we aim to develop an automated FoG detector for subject independent. After extracting highly relevant features, we apply anomaly detection techniques to detect FoG events. Specifically, feature selection is performed using correlation and clusterability metrics. From a list of 244 feature candidates, 36 candidates were selected using saliency and robustness criteria. We develop an anomaly score detector with adaptive thresholding to identify FoG events. Then, using accuracy metrics, we reduce the feature list to seven candidates. Our novel multichannel freezing index was the most selective across all window sizes, achieving sensitivity (specificity) of (). On the other hand, freezing index from the vertical axis was the best choice for a single input, achieving sensitivity (specificity) of () for ankle and () for back sensors. Our subject-independent method is not only significantly more accurate than those previously reported, but also uses a much smaller window (e.g., versus ) and/or lower tolerance (e.g., versus ).Freezing of gait (FoG) is common in Parkinsonian gait and strongly relates to falls. Current clinical FoG assessments are patients' self-report diaries and experts' manual video analysis. Both are subjective and yield moderate reliability. Existing detection algorithms have been predominantly designed in subject-dependent settings. In this paper, we aim to develop an automated FoG detector for subject independent. After extracting highly relevant features, we apply anomaly detection techniques to detect FoG events. Specifically, feature selection is performed using correlation and clusterability metrics. From a list of 244 feature candidates, 36 candidates were selected using saliency and robustness criteria. We develop an anomaly score detector with adaptive thresholding to identify FoG events. Then, using accuracy metrics, we reduce the feature list to seven candidates. Our novel multichannel freezing index was the most selective across all window sizes, achieving sensitivity (specificity) of (). On the other hand, freezing index from the vertical axis was the best choice for a single input, achieving sensitivity (specificity) of () for ankle and () for back sensors. Our subject-independent method is not only significantly more accurate than those previously reported, but also uses a much smaller window (e.g., versus ) and/or lower tolerance (e.g., versus ).

Assessing the mean strength and variations of the time-to-time fluctuations of resting-state brain activity

The time-to-time fluctuations (TTFs) of resting-state brain activity as captured by resting-state fMRI (rsfMRI) have been repeatedly shown to be informative of functional brain structures and disease-related alterations. TTFs can be characterized by the mean and the range of successive difference. The former can be measured with the mean squared successive difference (MSSD), which is mathematically similar to standard deviation; the latter can be calculated by the variability of the successive difference (VSD). The purpose of this study was to evaluate both the resting state-MSSD and VSD of rsfMRI regarding their test-retest stability, sensitivity to brain state change, as well as their biological meanings. We hypothesized that MSSD and VSD are reliable in resting brain; both measures are sensitive to brain state changes such as eyes-open compared to eyes-closed condition; both are predictive of age. These hypotheses were tested with three rsfMRI datasets and proven true, suggesting both MSSD and VSD as reliable and useful tools for resting-state studies.

Functional specificity of rat vibrissal primary afferents

In this study, we propose to analyze the peripheral vibrissal system specificity through its neuronal responses. Receiver operating characteristics (ROC) curve analyses were used, which required the implementation of a binary classifier (artificial neural network) trained to identify the applied stimulus. The training phase consisted of the observation of a predetermined amount of vibrissal sweeps on two surfaces of different texture and similar roughness. Our results suggest that the specificity of the peripheral vibrissal system easily permits the discrimination between perceived stimuli, quantified through neuronal responses, and that it can be evaluated through an ROC curve analysis. We found that such specificity makes a linear binary classifier capable of detecting differences between stimuli with five sweeps at most.

Design of a novel biomedical signal processing and analysis tool for functional neuroimaging

In this paper, a MATLAB-based graphical user interface (GUI) software tool for general biomedical signal processing and analysis of functional neuroimaging data is introduced. Specifically, electroencephalography (EEG) and electrocardiography (ECG) signals can be processed and analyzed by the developed tool, which incorporates commonly used temporal and frequency analysis methods. In addition to common methods, the tool also provides non-linear chaos analysis with Lyapunov exponents and entropies; multivariate analysis with principal and independent component analyses; and pattern classification with discriminant analysis. This tool can also be utilized for training in biomedical engineering education. This easy-to-use and easy-to-learn, intuitive tool is described in detail in this paper.

Relating structure and function in the human brain: relative contributions of anatomy, stationary dynamics, and non-stationarities

Investigating the relationship between brain structure and function is a central endeavor for neuroscience research. Yet, the mechanisms shaping this relationship largely remain to be elucidated and are highly debated. In particular, the existence and relative contributions of anatomical constraints and dynamical physiological mechanisms of different types remain to be established. We addressed this issue by systematically comparing functional connectivity (FC) from resting-state functional magnetic resonance imaging data with simulations from increasingly complex computational models, and by manipulating anatomical connectivity obtained from fiber tractography based on diffusion-weighted imaging. We hypothesized that FC reflects the interplay of at least three types of components: (i) a backbone of anatomical connectivity, (ii) a stationary dynamical regime directly driven by the underlying anatomy, and (iii) other stationary and non-stationary dynamics not directly related to the anatomy. We showed that anatomical connectivity alone accounts for up to 15% of FC variance; that there is a stationary regime accounting for up to an additional 20% of variance and that this regime can be associated to a stationary FC; that a simple stationary model of FC better explains FC than more complex models; and that there is a large remaining variance (around 65%), which must contain the non-stationarities of FC evidenced in the literature. We also show that homotopic connections across cerebral hemispheres, which are typically improperly estimated, play a strong role in shaping all aspects of FC, notably indirect connections and the topographic organization of brain networks.

Fusion during entrainment of orthodromic reciprocating tachycardia is enhanced for basal pacing sites but diminished when pacing near Purkinje system end points

BACKGROUND

In the electrophysiological laboratory, orthodromic atrioventricular reciprocating tachycardia (ORT) can be distinguished from atrial tachycardia and atrioventricular node reentry tachycardia by identifying orthodromic and antidromic wavefront fusion during ventricular overdrive pacing (VOP). Previous work has shown that basal VOP near the accessory pathway (AP) increases the likelihood of observing fusion; however, in a third of cases, fusion is not appreciable regardless of VOP location.

OBJECTIVE

To explore the hypothesis that pacing near His-Purkinje system (PS) end points reduces fusion quality, which may explain patients with nonresponsive ORT.

METHODS

In a novel computer model of ORT, simulations were performed with a variety of AP locations and pacing sites; results were analyzed to assess factors influencing fusion quality in pseudo-electrocardiogram signals.

RESULTS

Entrainment by basal VOP near the AP was more likely to produce fusion visible on simulated electrocardiograms compared to entrainment by apical VOP, but this advantage was dramatically diminished when the pacing site was also near PS end points. Prediction of fusion quality based on AP proximity alone was dramatically improved when corrected to penalize for PS proximity.

CONCLUSIONS

These results suggest that basal VOP near the AP and far from the PS is optimal; this could be tested in patients. A denser basal ramification of PS fibers is known to exist in a minority of human hearts; our findings indicate that this unusual PS configuration is a plausible explanation for ORT cases where fusion is never observed in spite of entrainment by basal VOP near the AP.

Ultrasound speed varies in articular cartilage under indentation loading

In ultrasound elastography, tissue strains are determined by localizing changes in ultrasound echoes during mechanical loading. The technique has been proposed for arthroscopic quantification of the mechanical properties of cartilage. The accuracy of ultrasound elastography depends on the invariability of sound speed in loaded tissue. In unconfined geometry, mechanical compression has been shown to induce variation in sound speed, leading to errors in the determined mechanical properties. This phenomenon has not been confirmed in indentation geometry, the only loading geometry applicable in situ or in vivo. In the present study, ultrasound speed during indentation of articular cartilage was characterized and the effect of variable sound speed on the strain measurements was investigated. Osteochondral samples (n = 7, diameter = 25.4 mm), prepared from visually intact bovine patellae (n = 7), were indented with a plane-ended ultrasound transducer (diameter = 5.6 mm, peak frequency: 8.1 MHz). A sequence of three compression tests (strain-rate = 10%/s, 2700-s relaxation) was applied using the mean strains of 2.2%, 4.5%, and 6.4%. Then, ultrasound speed during the ramp and stress-relaxation phases was determined using the time-of- flight technique. To investigate the role of cartilage structure and composition for sound speed in loaded articular cartilage, a sample-specific fibril-reinforced poroviscoelastic (FRPVE) finite element model was constructed and fitted to experimental mechanical data. Ultrasound speed in articular cartilage decreased significantly during dynamic indentation (p <; 0.05). The magnitude of the decrease in speed during indentation was related to the applied strain. However, the relative error in acoustically determined tissue strain was inversely related to the magnitude of true strain. The modeling results suggested that the compression-related variation in sound speed is controlled by changes in the collagen architecture during dynamic indentation. To conclude, variation in sound speed during dynamic indentation of articular cartilage may lead to significant errors in the values of measured mechanical parameters. Because the relative errors are inversely proportional to applied strain, higher strains should be used to minimize the errors in, e.g., in vivo measurements.

Improving ECG classification accuracy using an ensemble of neural network modules

This paper illustrates the use of a combined neural network model based on Stacked Generalization method for classification of electrocardiogram (ECG) beats. In conventional Stacked Generalization method, the combiner learns to map the base classifiers' outputs to the target data. We claim adding the input pattern to the base classifiers' outputs helps the combiner to obtain knowledge about the input space and as the result, performs better on the same task. Experimental results support our claim that the additional knowledge according to the input space, improves the performance of the proposed method which is called Modified Stacked Generalization. In particular, for classification of 14966 ECG beats that were not previously seen during training phase, the Modified Stacked Generalization method reduced the error rate for 12.41% in comparison with the best of ten popular classifier fusion methods including Max, Min, Average, Product, Majority Voting, Borda Count, Decision Templates, Weighted Averaging based on Particle Swarm Optimization and Stacked Generalization.

Comparison of ultrasound speed in articular cartilage measured by different time-of-flight methods

The purpose of this study was to investigate whether different time-of-flight (TOF) methods including amplitude-related methods, which determine tissue borders from the reflected wave itself, and the cross-correlation method, which requires reference signals to determine borders, influence speed of sound (SOS) values for articular cartilage. Left femoral condyle samples from a 6-month-old pig and a 3-year-old pig were used. Radiofrequency signals from the cartilage surface and cartilage-bone interface were acquired using the ultrasound transducer for nine points in each sample. TOF was calculated by three amplitude-related methods (peak amplitude, peak envelope, signal phase) and a cross-correlation method. Cartilage thickness was measured microscopically, and SOS was calculated at each point. Mean (± standard deviation) SOSs in cartilage from the 9-point measurement by the four TOF methods were 1488 ± 51, 1488 ± 48, 1487 ± 54, and 1466 ± 51 m/s (for peak amplitude, peak envelope, signal phase, and cross-correlation methods, respectively) for the 6-month-old pig, and 1709 ± 107, 1717 ± 104, 1713 ± 105, and 1695 ± 138 m/s, respectively, for the 3-year-old pig. Paired t testing identified no significant differences between the amplitude-related methods and the cross-correlation method, although SOS values yielded by the amplitude-related methods tended to be higher than those from the cross-correlation method. These results suggest that amplitude-related methods of TOF measurement and the cross-correlation method are equivalently applicable to articular cartilage SOS measurement when a wave is clearly reflected from cartilage. TOF methods should thus be considered in studies on SOS measurement.

Brain tissue oxygen amperometry in behaving rats demonstrates functional dissociation of dorsal and ventral hippocampus during spatial processing and anxiety

Traditionally, the function of the hippocampus (HPC) has been viewed in unitary terms, but there is growing evidence that the HPC is functionally differentiated along its septotemporal axis. Lesion studies in rodents and functional brain imaging in humans suggest a preferential role for the septal HPC in spatial learning and a preferential role for the temporal HPC in anxiety. To better enable cross-species comparison, we present an in vivo amperometric technique that measures changes in brain tissue oxygen at high temporal resolution in freely-moving rats. We recorded simultaneously from the dorsal (septal; dHPC) and ventral (temporal; vHPC) HPC during two anxiety tasks and two spatial tasks on the radial maze. We found a double-dissociation of function in the HPC, with increased vHPC signals during anxiety and increased dHPC signals during spatial processing. In addition, dHPC signals were modulated by spatial memory demands. These results add a new dimension to the growing consensus for a differentiation of HPC function, and highlight tissue oxygen amperometry as a valuable tool to aid translation between animal and human research.

Evaluation of EMG processing techniques using Information Theory

Background

Electromyographic signals can be used in biomedical engineering and/or rehabilitation field, as potential sources of control for prosthetics and orthotics. In such applications, digital processing techniques are necessary to follow efficient and effectively the changes in the physiological characteristics produced by a muscular contraction. In this paper, two methods based on information theory are proposed to evaluate the processing techniques.

Methods

These methods determine the amount of information that a processing technique is able to extract from EMG signals. The processing techniques evaluated with these methods were: absolute mean value (AMV), RMS values, variance values (VAR) and difference absolute mean value (DAMV). EMG signals from the middle deltoid during abduction and adduction movement of the arm in the scapular plane was registered, for static and dynamic contractions. The optimal window length (segmentation), abduction and adduction movements and inter-electrode distance were also analyzed.

Results

Using the optimal segmentation (200 ms and 300 ms in static and dynamic contractions, respectively) the best processing techniques were: RMS, AMV and VAR in static contractions, and only the RMS in dynamic contractions. Using the RMS of EMG signal, variations in the amount of information between the abduction and adduction movements were observed.

Conclusions

Although the evaluation methods proposed here were applied to standard processing techniques, these methods can also be considered as alternatives tools to evaluate new processing techniques in different areas of electrophysiology.

Synchronization in the bivariate intrauterine pressure signals' nonlinear dynamics methods

It was shown earlier that nonlinear processes probably generate uterine contractions. In this study, the nonlinear synchronization measures (the mutual correlation dimension, the cross-approximate entropy, the mutual information and the nonlinear interdependencies) were employed to analyze the association between two time series representing the uterine contraction activity. Here the notion of synchronization is used in a loose sense as the synonym of correlation, the similarity of the signals or the similarity of their dynamics. The signals were recorded from the different topographic regions of the uterus: the cervix and the fundus. The results obtained by means of different algorithms are different but qualitatively similar for the checked methods.

Differences in acoustic properties of intact and degenerated human patellar cartilage during compression

Ultrasound indentation measurements have been shown to provide means to assess cartilage integrity and mechanical properties. To determine cartilage stiffness in the ultrasound indentation geometry, cartilage is compressed with an ultrasound transducer to determine the induced strain from the ultrasound signal using the time-of-flight principle. As the ultrasound speed in cartilage has been shown to vary during compression, the assumption of constant speed generates significant errors in the values of mechanical parameters. This variation in ultrasound speed has been investigated in intact cartilage, however, its existence and significance in degenerated tissue is unknown. In the present study, we investigate this issue with both intact and spontaneously degenerated human tissue. To accomplish this aim, we determined ultrasound speed and attenuation in human patellar cartilage (n=68) during mechanical loading. For reference, cartilage mechanical properties and proteoglycan, collagen and water contents were determined. The acoustic properties were related to the composition and mechanical properties of the samples. Ultrasound speed showed significant, site-dependent variation and it was significantly associated (r=0.79-0.81, p<0.01) with the mechanical properties of cartilage. The compression related decrease in ultrasound speed showed statistically significant variation between different stages of degeneration. Error simulations revealed that changes in ultrasound speed during 2% compression could generate errors up to 15% in the values of elastic moduli of samples with early degeneration, if determined with the ultrasound indentation technique. In samples with advanced degeneration, the error was significantly (p<0.05) smaller being 2% on average. As the compression related variation in ultrasound speed was lower in more degenerated samples, the mechanical parameters could be diagnosed more reliably in tissue showing advanced degeneration. The present results address the need to consider possible uncertainties in mechano-acoustic measurements of articular cartilage and call for methods to correct the effect of variable sound speed during compression.

Determination of low-pass filter cutoff frequencies for high-rate biomechanical signals obtained using videographic analysis

Diffuse brain injury (DBI) commonly results from blunt impact followed by sudden head rotation, wherein severity is a function of rotational kinematics. A noninvasive in vivo rat model was designed to further investigate this relationship. Due to brain mass differences between rats and humans, rotational acceleration magnitude indicative of rat DBI ( approximately 350 krad/s(2)) has been estimated as approximately 60 times greater than that of human DBI ( approximately 6 krad/s(2)). Prior experimental testing attempted to use standard transducers such as linear accelerometers to measure loading kinematics. However, such measurement techniques were intrusive to experimental model operation. Therefore, initial studies using this experimental model obtained rotational displacement data from videographic images and implemented a finite difference differentiation (FDD) method to obtain rotational velocity and acceleration. Unfortunately, this method amplified high-frequency, low-amplitude noise, which interfered with signal magnitude representation. Therefore, a coherent average technique was implemented to improve the measurement of rotational kinematics from videographic images, and its results were compared with those of the previous FDD method. Results demonstrated that the coherent method accurately determined a low-pass filter cutoff frequency specific to pulse characteristics. Furthermore, noise interference and signal attenuation were minimized compared with the FDD technique.

Strain-dependent modulation of ultrasound speed in articular cartilage under dynamic compression

Mechanical properties of articular cartilage may be determined by means of mechano-acoustic indentation, a clinically feasible technique for cartilage diagnostics. Unfortunately, ultrasound speed varies in articular cartilage during mechanical compression. This can cause significant errors to the measured mechanical parameters. In this study, the strain-dependent variation in ultrasound speed was investigated during dynamic compression. In addition, we estimated errors that were induced by the variation in ultrasound speed on the mechano-acoustically measured elastic properties of the tissue. Further, we validated a computational method to correct these errors. Bovine patellar cartilage samples (n = 7) were tested under unconfined compression. Strain-dependence of ultrasound speed was determined under different compressive strains using an identical strain-rate. In addition, the modulation of ultrasound speed was simulated using the transient compositional and structural changes derived from fibril-reinforced poroviscoelastic (FRPVE) model. Experimentally, instantaneous compressive strain modulated the ultrasound speed (p < 0.05) significantly. The decrease of ultrasound speed was found to change nonlinearly as a function of strain. Immediately after the ramp loading ultrasound speed was found to be changed -0.94%, -1.49%, -1.84%, -1.87%, -1.89% and -2.15% at the strains of 2.4%, 4.9%, 7.3%, 9.7%, 12.1% and 14.4%, respectively. The numerical simulation revealed that the compression-related decrease in ultrasound speed induces significant errors in the mechano-acoustically determined strain (39.7%) and dynamic modulus (72.1%) at small strains, e.g., at 2.4%. However, at higher strains, e.g., at 14.4%, the errors were smaller, i.e., 12.6% for strain and 14.5% for modulus. After the proposed computational correction, errors related to ultrasound speed were decreased. By using the correction, with e.g., 2.4% strain, errors in strain and modulus were decreased from 39.7% to 7.2% and from 72.1% to 35.3%, respectively. The FRPVE model, addressing the changes in fibril orientation and void ratio during compression, showed discrepancy of less than 1% between the predicted and measured ultrasound speed during the ramp compression.

Effects of ultrasound frequency, temporal sampling frequency, and spatial sampling step on the quantitative ultrasound parameters of articular cartilage

Quantitative ultrasound imaging may provide a technique for diagnosing initial signs of osteoarthritis (OA), such as surface fibrillation of articular cartilage. Because subchondral sclerosis and osteophyte formation occur in OA as well, ultrasonic analysis of subchondral bone could yield useful diagnostic information. In this study, we investigated whether low-frequency (5 MHz) ultrasound, typically used in bone diagnostics, would be feasible for evaluating the integrity of the surface of the cartilage. The reflection parameters in the time and frequency domains, the ultrasound roughness index, and the wavelet-based parameters were evaluated using ultrasound transducers operating at 5, 10, and 50 MHz frequencies. The effects of variable size of spatial sampling steps and of temporal sampling frequencies were also investigated. Custom-made phantoms and cartilage samples with various surface characteristics were analyzed. The reflection parameters detected the surface degradation with all ultrasound frequencies. The roughness of the surface could only be evaluated reliably with the 50 MHz-focused transducer. In conclusion, simultaneous analysis of the reflection parameters of the cartilage and the subchondral bone is feasible at low (5 MHz) ultrasound frequencies. However, reliable evaluation of the microtopography of the cartilage requires use of a higher ultrasound frequency.

Different subtypes of striatal neurons are selectively modulated by cortical oscillations

The striatum is the key site for cortical input to the basal ganglia. Cortical input to striatal microcircuits has been previously studied only in the context of one or two types of neurons. Here, we provide the first description of four putative types of striatal neurons (medium spiny, fast spiking, tonically active, and low-threshold spiking) in a single data set by separating extracellular recordings of sorted single spikes recorded under halothane anesthesia using waveform and burst parameters. Under halothane, the electrocorticograms and striatal local field potential displayed spontaneous oscillations at both low (2-9 Hz) and high (35-80 Hz) frequencies. Putative fast spiking interneurons were significantly more likely to phase lock to high-frequency cortical oscillations and displayed significant cross-correlations in this frequency range. These findings suggest that, as in neocortex and hippocampus, the coordinated activity of fast spiking interneurons may specifically be involved in mediating oscillatory synchronization in the striatum.

Effect of bone marrow on acoustic properties of trabecular bone--3D finite difference modeling study

The composition of bone marrow is influenced by many factors, such as age and diseases. The present numerical study investigates the contribution of marrow on the acoustic measurements of trabecular bone. Cylindrical bone samples (n = 11), extracted from three anatomical sites of human cadaver knees, were imaged with a high-resolution microtomography (microCT). Three-dimensional finite difference time domain (FDTD) models (Wave 3000 Pro 2.2, Cyberlogic Inc., NY, USA) were created using the segmented microCT images of each sample. First, we evaluated the effect of voxel size on the computer resource requirements, morphological parameters and acoustic simulations. Second, the effect of bone marrow on ultrasonic measurements was assessed. The simulations were repeated with two voxel sizes before and after substitution of bone marrow (i.e., fat) with water. The voxel size of the FDTD mesh controlled the fine structure of the modeled calcified matrix and significantly affected the simulation results. However, present simulations showed that the effect of bone marrow on ultrasound parameters can be reliably simulated with the applied voxel sizes of 72 and 90 microm. Ultrasound attenuation and speed were found (p < 0.01) to decrease and increase, respectively, when bone marrow was substituted with water. Moreover, reflection from the surface of the sample increased (p < 0.01) and backscatter from internal structures decreased (p < 0.01) after removal of marrow. The effect of bone marrow on the acoustic properties was stronger in samples with low bone volume fraction. The present results indicate that the amount and quality of bone marrow significantly influence the acoustic properties of trabecular bone. Possible interindividual differences in the composition of bone marrow may increase uncertainty in clinical ultrasound diagnostics of osteoporosis. Importantly, the effect is most significant in osteoporotic low-density bone.

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.

Characterization of T wave alternans with ambulatory electrocardiography

T wave alternans (TWA) is a marker of ventricular electrical instability considered to be predictive for ventricular tachyarrhythmias. Techniques have been developed to detect TWA at the microv level as a method for arrhythmia risk stratification of persons at high risk for sudden cardiac death. Currently, TWA is typically calculated using spectral analysis, whereby TWA is presumed to assume characteristics of stationarity. In contrast, a nonspectral method known as modified moving average analysis is purported to detect transient TWA that would not be observed using a spectral approach. The purpose of this pilot study work was to establish the basic TWA signal properties obtained with a device developed by GE Medical Systems using a descriptive, correlational study design. Ambulatory electrocardiography (AECG) recordings (N = 24) were digitized and processed, and TWA was calculated via the modified moving average technique. Findings showed that noise was positively correlated with TWA in AECG channel 1 (r = .899, p < .01) and AECG channel 2 (r = .758, p < .01). However, no significant difference (p = .237) was observed in TWA values between the AECG channels. A weak positive correlation was found between TWA and heart rate, expressed as beats per min (r = .262). Heart rate mildly predicted TWA (R = 0.34). Nonstationarity was evaluated by testing for trend and randomness. TWA values measured from AECG recordings were found to be influenced moderately by noise and minimally by heart rate and lead placement.

Ultrasound speed in articular cartilage under mechanical compression

Ultrasound elastography is a method that can be used to determine the elastic properties of soft tissues and it has been recently applied to study of articular cartilage. While ultrasound elastography techniques assume a constant ultrasound speed in tissue under mechanical compression, ultrasound speed in articular cartilage has been found to vary depending on the loading conditions. This may limit the quantitative use of the technique for determination of the elastic properties of articular cartilage along the axis of ultrasound propagation. The aim of the present study was to investigate the origin of the load-related variation in ultrasound speed. Samples of human and bovine articular cartilage (n = 82) were mechanically and acoustically tested during unconfined compression. A statistically significant (p < 0.05) variation of ultrasound speed was found in cartilage during a stress-relaxation test. A finite element model was constructed by exploiting microscopically determined collagen and proteoglycan contents, collagen orientation and biochemical analyses of water content. From the finite element simulations, collagen orientation and the void ratio (fluid-to-solid ratio) as a function of time were assessed and, together with the experimentally determined ultrasound speed, a linear model predicting variation of the ultrasound speed in human articular cartilage under mechanical compression was established. The model predicted compression-related ultrasound speed with an error of <0.3% at each time point. The effect of strain rate on variation of ultrasound speed was tested in bovine cartilage samples. The decrease in ultrasound speed was found to be proportional to the strain rate. The results suggest that ultrasound speed in articular cartilage is controlled mainly by collagen orientation and the void ratio and depends on the imposed strain rate. A numerical simulation revealed that the compression-related decrease in ultrasound speed induces notable errors in mechano-acoustically determined strain. A method to eliminate the compression-related errors in measured strain and elastic properties may be needed in mechano-acoustic measurements of articular cartilage.

Differences in the characteristics of induced and spontaneous episodes of ventricular fibrillation

AIMS

The degree of organization of ventricular fibrillation (VF) can be examined in terms of the regularity of the electrical activity within the ventricle. Using electrograms (EGMs) stored within implanted cardioverter defibrillators (ICDs), we examined the hypothesis that the degree of organization, or regularity, was different if the VF was induced by electrical stimulation as opposed to occurring clinically due to ischemia or scar.

METHODS AND RESULTS

We compared the statistical characteristics of EGMs recorded by ICDs during spontaneous episodes with those induced during device testing in the laboratory in nine subjects. Regularity of the VF EGM signals was quantified using autocorrelation, Shannon entropy (derived from cycle to cycle activation complexes), and Kolmogorov entropy (derived from eight second long episodes of VF). All three measurements showed a statistically greater degree of regularity for induced VF than in spontaneous episodes.

CONCLUSION

Analysis of VF EGMs using these techniques is novel and robust, providing a new way for assessing electrical organization during VF. The clinical significance and utility of differences in VF waveform regularity is unclear at this stage.

ECG Beat Detection Using a Geometrical Matching Approach

In the framework of the electrocardiography (ECG) signals, this paper describes an original approach to identify heartbeat morphologies and to detect R-wave events. The proposed approach is based on a "geometrical matching" rule evaluated using a decision function in a local moving-window procedure. The decision function is a normalized measurement of a similarity criterion comparing the windowed input signal with the reference beat-pattern into a nonlinear-curve space. A polynomial expansion model describes the reference pattern. For the curve space, an algebraic-fitting distance is built according to the canonical equation of the unit circle. The geometrical matching approach operates in two stages, i.e., training and detection ones. In the first stage, a learning-method based on genetic algorithms allows us estimating the decision function from training beat-pattern. In the second stage, a level-detection algorithm evaluates the decision function to establish the threshold of similarity between the reference pattern and the input signal. Finally, the findings for the MIT-BIH Arrhythmia Database present about 98% of sensitivity and 99% of positive predictivity for the R-waves detection, using low-order polynomial models.

Blood pressure response to isometric exercise in patients with peripheral atherosclerotic disease

BACKGROUND

The purpose of this study was to compare the circulatory responses to isometric exercise in patients with peripheral atherosclerotic disease (PAD) with healthy controls.

METHODS

Eleven patients with diagnosed PAD, a control group of eleven healthy young adults, and a control group of eleven healthy age-matched adults participated. Blood pressure, heart rate, stroke volume, cardiac output, blood velocity in the brachial artery, acral skin perfusion was continuously recorded and total peripheral resistance calculated before, during and after 2 min of 40% maximum voluntary contraction of the forearm.

RESULTS

At rest we found a consistently higher level of mean arterial pressure (MAP) and systolic pressure (SP) in the elderly, both PAD patients and elderly controls, compared with the young controls. We found no significant difference in diastolic blood pressure. Two minutes isometric handgrip exercise induced a similar increase in MAP in all three groups (patients 32.6 (17.9) mm Hg [mean (SD)], young control group 25.3 (8.9) mm Hg, age-matched control group 36.1 (10.6) mm Hg). No significant differences were found in the other measured cardiovascular variables during isometric handgrip. Increased TPR is the main factor contributing to the increase in blood pressure in all three groups.

CONCLUSION

Our study indicates that the pressor response continues to be well regulated with age, also when the cardiovascular system is altered by marked atherosclerosis. The consequence is that both PAD patients and elderly controls reach higher SP values during isometric exercise due to higher SP baseline values.

Uterine Contractility: Visualization of Synchronization Measures in Two Simultaneously Recorded Signals

The analysis of the uterine contraction signals in nonpregnant states gives information about physiological changes during the menstrual cycle. Spontaneous uterine activity was recorded directly by a dual microtip catheter. The device consisted of two ultra-miniature pressure sensors. One sensor was placed in the fundus, the other in the cervix. It was important to identify time delays between contractions in two topographic locations, which may be of potential diagnostic significance in various pathologies: dysmenorrhea, endometriosis, and fecundity disorders. In this study the following synchronization measures-the cross-correlation, the semblance, the mutual information-were used to visualize the time delay changes over time. These measures were computed in a moving window with a width corresponding to approximately two or three contractions. As a result, the running synchronization functions were obtained. The running synchronization functions visualize changes in the propagation of the two simultaneously recorded signals. The propagation% parameter assessed from these functions allows for quantitative description of synchronization. Finally, we illustrate the use of running synchronization functions to investigate the effect of treatment with tamoxifen on primary dysmenorrhea.

Effect of short-term, low-dose treatment with tamoxifen in patients with primary dysmenorrhea

Current treatment of painful periods and other symptoms related to primary dysmenorrhea (PD) is usually commenced with non-steroidal anti-inflammatory drugs or oral contraceptives, which fails in about 10% of affected patients. Tamoxifen, a selective estrogen-receptor modulator (SERM), has been demonstrated to directly inhibit uterine contractions, causing improvement in uterine blood flow. It could be considered for application in selected groups of dysmenorrheic patients, for instance carriers of breast cancer-associated antigen (BRCA) genes, breast cancer survivors or women with advanced endometriosis. Thus the aim of the present study was to investigate the effect of short-term treatment with tamoxifen on PD and PD-related symptoms, as well as its direct effect on parameters of intrauterine pressure during the painful menstruation, in a group of dysmenorrheic patients. After two cycles of administration of tamoxifen we noted a significant decrease in bleeding together with reductions in the severity of menstrual cramps, diarrhea, headache, fatigue and anxiety. In intrauterine pressure assessments, tamoxifen significantly decreased propagation of uterine contractions. In conclusion, SERMs such as tamoxifen may constitute a therapeutic option in selected groups of patients, improving dysmenorrheic symptoms. Additionally to its receptor-mediated effects, tamoxifen was shown to exert a direct influence on uterine contractile activity that may explain the decrease of menstrual pain and cramps noted in the studied group.

Redundancy of Independent Component Analysis in Four Common Types of Childhood Epileptic Seizure

SUMMARY

Independent component analysis (ICA) has recently been applied to epileptic seizure in the EEG. In this paper, the authors show how the fundamental axioms required for ICA to be valid are broken. Four common cases of childhood seizure are presented and assessed for stationarity and an eigenvalue analysis is applied. In all cases, for the stationary sections of data the eigenvalue analysis yields results that imply the signals are coming from a source-rich environment, thus yielding ICA inappropriate when applied to the four common types of childhood seizure. The results suggest that it is not appropriate to apply ICA or source localization from independent components in these four common cases of epilepsy, because the spurious independent components determined by ICA could lead to a spurious localization of the epilepsy. If surgery were to follow, it could result in the incorrect treatment of a healthy localized region of the brain.

Texture discrimination and multi-unit recording in the rat vibrissal nerve

Background

Rats distinguish objects differing in surface texture by actively moving their vibrissae. In this paper we characterized some aspects of texture sensing in anesthetized rats during active touch. We analyzed the multifiber discharge from a deep vibrissal nerve when the vibrissa sweeps materials (wood, metal, acrylic, sandpaper) having different textures. We polished these surfaces with sandpaper (P1000) to obtain close degrees of roughness and we induced vibrissal movement with two-branch facial nerve stimulation. We also consider the change in pressure against the vibrissa as a way to improve the tactile information acquisition. The signals were compared with a reference signal (control) – vibrissa sweeping the air – and were analyzed with the Root Mean Square (RMS) and the Power Spectrum Density (PSD).

Results

We extracted the information about texture discrimination hidden in the population activity of one vibrissa innervation, using the RMS values and the PSD. The pressure level 3 produced the best differentiation for RMS values and it could represent the "optimum" vibrissal pressure for texture discrimination. The frequency analysis (PSD) provided information only at low-pressure levels and showed that the differences are not related to the roughness of the materials but could be related to other texture parameters.

Conclusion

Our results suggest that the physical properties of different materials could be transduced by the trigeminal sensory system of rats, as are shown by amplitude and frequency changes. Likewise, varying the pressure could represent a behavioral strategy that improves the information acquisition for texture discrimination.

Visualization of synchronization of the uterine contraction signals: Running cross-correlation and wavelet running cross-correlation methods

In physiological research, we often study multivariate data sets, containing two or more simultaneously recorded time series. The aim of this paper is to present the cross-correlation and the wavelet cross-correlation methods to assess synchronization between contractions in different topographic regions of the uterus. From a medical point of view, it is important to identify time delays between contractions, which may be of potential diagnostic significance in various pathologies. The cross-correlation was computed in a moving window with a width corresponding to approximately two or three contractions. As a result, the running cross-correlation function was obtained. The propagation% parameter assessed from this function allows quantitative description of synchronization in bivariate time series. In general, the uterine contraction signals are very complicated. Wavelet transforms provide insight into the structure of the time series at various frequencies (scales). To show the changes of the propagation% parameter along scales, a wavelet running cross-correlation was used. At first, the continuous wavelet transforms as the uterine contraction signals were received and afterwards, a running cross-correlation analysis was conducted for each pair of transformed time series. The findings show that running functions are very useful in the analysis of uterine contractions.

A similarity measure for case based reasoning modeling with temporal abstraction based on cross-correlation

Adverse drug events (ADEs) are a major limitation of drug safety. They are often caused by inappropriate selection of dose and the concurrent use of drugs modulating each other (drug interaction). Risk assessment and prevention strategies must therefore consider co-administered drugs, individual doses, and their timing. In a new approach we evaluated the performance of cross correlation, commonly used in signal processing, to determine similarities in patient treatments. To achieve this, patient treatments were modeled as groups of vectors representing discrete time intervals. These vectors were cross-correlated and the results evaluated to find clusters in time courses indicating similarity in treatment of different patients. To evaluate our algorithm, we then created a number of test cases. The focus of this article is on each treatment, and its pattern in time and dosage. The algorithm successfully produces a relatively low similarity score for cases that are completely different with respect to their pattern of time and dosage but high scores when they are equal (score of 0.699) or similar (score of 0.528) in their therapies, and thus succeeds in having a relatively high specificity (27/30). Such an approach might help to considerably reduce the problem of false alarms which hampers most existing alerting systems for medication errors or impending ADEs.

Using data augmentation to obtain standard errors and conduct hypothesis tests in latent class and latent transition analysis

Latent class analysis (LCA) provides a means of identifying a mixture of subgroups in a population measured by multiple categorical indicators. Latent transition analysis (LTA) is a type of LCA that facilitates addressing research questions concerning stage-sequential change over time in longitudinal data. Both approaches have been used with increasing frequency in the social sciences. The objective of this article is to illustrate data augmentation (DA), a Markov chain Monte Carlo procedure that can be used to obtain parameter estimates and standard errors for LCA and LTA models. By use of DA it is possible to construct hypothesis tests concerning not only standard model parameters but also combinations of parameters, affording tremendous flexibility. DA is demonstrated with an example involving tests of ethnic differences, gender differences, and an Ethnicity x Gender interaction in the development of adolescent problem behavior.

Comparison of Trend Detection Algorithms in the Analysis of Physiological Time-Series Data

This paper presents a comparative performance analysis of various trend detection methods developed using fuzzy logic, statistical, regression, and wavelet techniques. The main contribution of this paper is the introduction of a new method that uses noise rejection fuzzy clustering to enhance the performance of trend detection methodologies. Furthermore, another contribution of this work is a comparative investigation that produced systematic guidelines for the selection of a proper trend detection method for different application requirements. Examples of representative physiological variables considered in this paper to examine the trend detection algorithms are: 1) blood pressure signals (diastolic and systolic); and 2) heartbeat rate based on RR intervals of electrocardiography signal. Furthermore, synthetic physiological data intentionally contaminated with various types of real-life noise has been generated and used to test the performance of trend detection methods and develop noise-insensitive trend-detection algorithms.

Effects of RR segment duration on HRV spectrum estimation

Although patterns of heart rate variability (HRV) hold considerable promise for clarifying issues in clinical applications, the inappropriate quantification and interpretation of these patterns may obscure critical issues or relationships and may impede rather than foster the development of clinical applications. The duration of the RR interval series is not a matter of convenience but a fine balance between two important issues: acceptable variance and stationarity of the time series on one hand, and acceptable resolution of the spectral estimate and reduced spectral leakage on the other. Further, in the standard short-term HRV analysis, it has been observed that the previous studies in HRV spectral analysis use a wide range of RR interval segment duration for spectral estimation by Welch's algorithm. The standardization of RR interval segment duration is also important for comparisons among studies and is essential for within-study experimental contrasts. In the present study, a comparative analysis for RR interval segment durations has been made to propose an optimal RR interval segment duration. Firstly a simulated signal was analyzed with Hann window and zero padding for the segment lengths of 1024, 512, 256 and 128 samples resampled at 4 Hz with 50% overlapping. Again, the above procedure was applied to RR interval series and it was concluded that segment length of 256 samples with 50% overlapping provides a smoothed spectral estimate with clearly outlined peaks in low- and high-frequency bands. This easily understandable and interpretable spectral estimate leads to a better visual and automated analysis, which is not only desirable in basic physiology studies, but also a prerequisite for a widespread utilization of frequency domain techniques in clinical studies, where simplicity and effectiveness of information are of primary importance.

Sleep-dependent changes in the coupling between heart period and arterial pressure in newborn lambs

This study assessed whether sleep-dependent changes in the relationship between heart period (HP) and mean arterial pressure (MAP) occur in newborn life. Electrodes for electrocorticographic, electromyographic, and electrooculographic monitoring and an arterial catheter for blood pressure recordings were implanted in 11 newborn lambs. HP and MAP beat-to-beat values were computed from 120-s blood pressure recordings during quiet wakefulness, active sleep, and quiet sleep. For each recording, the time shift at which the maximum of the HP versus MAP cross-correlation function was attained was identified. For each lamb and wake-sleep state, an average correlation coefficient was then computed corresponding to the median value of such time shifts. The maximum of the cross-correlation function was attained with HP lagging behind MAP. The corresponding mean correlation coefficient was significantly higher in quiet sleep (0.51 +/- 0.05) than either in quiet wakefulness (0.31 +/- 0.05) or in active sleep (0.29 +/- 0.03). Sleep-related differences in the correlation between HP and MAP were maintained after HP and MAP data were low-pass filtered at 0.3 Hz to remove their fast ventilatory oscillations. In conclusion, data indicate that the relationship between spontaneous fluctuations in HP and those in MAP is sleep-state dependent in newborn lambs. A positive HP versus MAP correlation with HP lagging behind MAP is consistent with baroreflex control of HP. Heart rhythm thus may be more tightly controlled by the baroreceptor reflex and less dependent on central autonomic commands in quiet sleep than either in quiet wakefulness or in active sleep.

A laser micromachined probe for recording multiple field potentials in the thalamus

Multichannel recording provides integral information about electrical brain activities at one instant in time. In this study, multielectrode probes were fabricated to record the thalamic field potentials (FPs) responding to the electrical stimulation of nerve at the rat tail. At first, the number of sweeps used to form the evoked FP average and the spatial sampling density were determined by using cross-correlation functions, which were then statistically analyzed. The difference was significant at P < 0.05, if the number of sweeps for averaging was more than 50 and the spatial interval between two consecutive recording sites was less than 50 microm in the anteroposterior, mediolateral and ventrodorsal directions. The responsive area was distributed vertically in the thalamus (ventral posterior lateral (VPL) nucleus); therefore, the recording sites were arranged in one linear array. Sixteen recording sites, which were 50 microm apart from each other, were distributed in the ventrodorsal direction. A 16-channel silicon probe was fabricated by using a standard photolithography process and laser micromachining techniques. The probe provides capabilities to record multiple thalamic evoked FPs and multiunit activities simultaneously.