Signal Standard Deviation Research Articles

Hard X-Ray Emission Detection Using Deep Learning Analysis of the Radiated UHF Electromagnetic Signal From a Plasma Focus Discharge

A method to determine the presence of hard X-ray emission processes from a dense plasma focus (205 J, 22 kV, 6.5 mbar H 2) using Ultra High Frequency (UHF) measurements and deep learning techniques is presented. Simultaneously, the electromagnetic UHF radiation emitted from the plasma focus was measured with a Vivaldi UHF antenna, while the hard X-ray emission was measured with a scintillator-photomultiplier system. A classification algorithm based on deep learning methods, using two-dimensional convolutional layers, was implemented to predict the hard X-ray signal standard deviation value using only the antenna signal measurement. Two independent datasets, consisting of 999 and 1761 data pairs each, were used in the analysis. Different realizations of the training/validation process using a deep learning model, obtained overall better results in comparison to other machine learning methods like $k$ -neighbors, decision trees, gradient boost, and random forest. The results of the deep learning algorithm, and even its comparison with other machine learning methods, indicate that a relationship between the electromagnetic UHF radiation and hard X-ray emission can be established, enabling the indirect detection of hard X-ray pulses only using the UHF antenna signal. This indirect detection presents the opportunity to have a simple and low-cost diagnostic, compared to the methods currently used to characterize the pulses of X-rays emitted from plasma focus discharges.

Drive and measurement electrode patterns for electrode impedance tomography (EIT) imaging of neural activity in peripheral nerve

Objective; To establish the performance of several drive and measurement patterns in EIT imaging of neural activity in peripheral nerve, which involves large impedance changes in the nerve’s anisotropic length axis. Approach; Twelve drive and measurement electrode patterns are compared using a finite element (FE) four-cylindrical shell model of a peripheral nerve and a 32 channel dual-ring nerve cuff. The central layer of the FE model contains impedance changes representative of neural activity of −0.30 in length axis and −8.8 × 10−4 in the radial axis. Six of the electrode patterns generate longitudinal drive current, which runs parallel to the anisotropic axis, while the remaining six patterns generate transverse drive current, which runs perpendicular to the anisotropic axis. Main results; Of the twelve patterns evaluated, transverse current patterns produce higher resolution than longitudinal current patterns but are also more susceptible to noise and errors, and exhibit poorer sensitivity to impedance changes in central sample locations. Three of the six longitudinal current patterns considered can reconstruct fascicle level impedance changes with up to 0.2 mV noise and error, which corresponds to between −5.5 and +0.18 dB of the normalised signal standard deviation. Reducing the spacing between the two electrode rings in all longitudinal current patterns reduced the signal to error ratio across all depth locations of the sample. Significance; Electrode patterns which target the large impedance change in the anisotropic length axis can provide improved robustness against noise and errors, which is a critical step towards real time EIT imaging of neural activity in peripheral nerve.

Wednesday, September 26, 2018 7:35 AM–9:00 AM ePosters: P42. Preliminary validation of a machine-learning MRI-based algorithm to identify patients with functional impairment associated with lumbar stenosis

BACKGROUND CONTEXT The anatomic changes characterized by magnetic resonance imaging (MRI) in patients with lumbar spinal stenosis weakly correlate with pain intensity and functional limitation in patients who endorse the classic evoked symptom pattern of neurogenic claudication (NC). PURPOSE To identify MRI features that correlate with the evoked symptom pattern of neurogenic claudication using a novel machine learning algorithm. STUDY DESIGN/SETTING Subjects with chronic low back pain and a lumbar MRI study were recruited from a single academic neurosurgery department and evaluated using a standardized protocol of validated, disease specific questionnaires and a treadmill-based functional assessment. PATIENT SAMPLE A total of 159 subjects with lumbar spinal stenosis were categorized into NC and nonNC cohorts based on clinical assessment. An independent radiologist blinded to clinical cohort analyzed the MRI data. Subjects with nonstandard MRI imaging data or studies compromised by artifact (eg lack of fat-suppression or presence of obscuring artifact) were excluded. 61 subjects (28 females) with 33 classified as NC+ were included. The mean age was 68.4±11. OUTCOME MEASURES Classification of NC based on imaging variables. METHODS The T2 weighted sagittal MRI FS images were segmented into posterior fat, subcutaneous fat, muscle, vertebral bodies, spinal canal and paravertebral space, then further segmented into seven sub-regions corresponding to levels T12 through S1. Image data was quantified for topography of each ROI or sub-ROI (area, compactness, and mass eccentricity); MRI signal (mean, standard deviation, skewness, kurtosis, entropy, and energy) and texture (gray-level co-occurrence matrix) (angular second moment, contrast, mutual information, correlation, dissimilarity, entropy, homogeneity, marginal entropy, and variance). Together the regions and measurements generated 1,512 features. Left and right mean and absolute difference were calculated to generate features representing asymmetry. All measurements were adjusted for gender and age differences; calibrated for gain differences using the characteristics of the background and normalized using the inverse rank transform. The normalized features were used to construct a diagnostic model based on logistic models. The model was constructed using a feature selection algorithm under a 10-fold cross-validation scheme repeated 5times. RESULTS Differences texture in the vertebral body and the spinal canal were strongly associated with the clinical diagnosis. The model validation predicts that MRI-based logistic models can separate NC with a sensitivity of 0.75 (95% CI 0.55–0.89) with a specificity of 0.97 (95% CI 0.84–1.00) and a diagnostic odds ratio of 96. CONCLUSIONS A machine learning model of MRI features can determine the likelihood of the clinical presentation of NC. Differences in vertebral body signal and the spinal canal show the strongest association to the clinical diagnosis. Further validation of the model may help facilitate pain management strategies.

Novel practical SNR determination method for MRI using double echo with longest second echo time (DELSET).

We devised a practical method using double echo with the longest second echo time (DELSET) for simple and accurate signal-to-noise ratio (SNR) measurement of MRIs. The DELSET method is based on the double-echo sequence in which the first and second echo times (TE) are set, respectively, as the clinically acceptable time for the signal image and as the longest time for the noise image. The second TE needs to be at least 8 times longer than T2 (for spin-echo) or T2* (for gradient-echo) of the objective tissue. For example, second TE > 560 ms for the case of T2 = 70 ms: the real part of signal intensity theoretically reaches the same order of magnitude as the quantization limit, due to the T2 relaxation process. SNR was calculated by dividing mean signal intensity in the first echo image by signal standard deviation (SD) in the second echo image in identical regions of interest after necessary noise correction. We determined the SNRs of cylindrical phantom images with different coils [quadrature (QD) and array coils] and sequences (spin-echo and spoiled gradient-echo sequences) and compared them between the DELSET and subtraction or background methods. The ratio of the mean signal intensity and SD in the second echo image with QD coil was determined to confirm whether the signal intensity had reached noise level. There were no significant differences in the phantom SNRs with both coils and sequences when DELSET was compared with the other two methods (p > 0.05 for all). The ratios of mean signal intensity and SD in the second echo images with both sequences were found to be in general agreement with a theoretical value. It was possible to obtain SNR images of the phantom, brain and abdomen with the DELSET method. The DELSET method enables simple and accurate SNR quantification. This practical method is applicable to in vivo parallel imaging. Advances in knowledge: Practical SNR quantification based on the DELSET method is feasible for application in MRI systems used clinically.

Computed diffusion weighted imaging (cDWI) and voxelwise-computed diffusion weighted imaging (vcDWI) for oncologic liver imaging: A pilot study

ObjectiveAim of the study was to evaluate the influence of the selection of measured b-values on the precision of cDWI in the upper abdomen as well as on the lesion contrast of PET-positive liver metastases in cDWI and vcDWI. MethodsWe performed a retrospective analysis of 10 patients (4 m, 63.5 ± 12.9 y/o) with PET-positive liver metastases examined in 3 T-PET/MRI with b = 100,600,800,1000 and 1500s/mm2. cDWI (cb1000/cb1500) and vcDWI were computed based on following combinations: i) b = 100/600 s/mm2, ii) b = 100/800 s/mm2, iii) b = 100/1000s/mm2, iv) b = 100/600/1000s/mm2 v) all measured b-values. Mean signal intensity (SI) and standard deviation (SD) in the liver, spleen, kidney, bone marrow and in liver lesions were acquired. The coefficient of variation (CV = SD/SI), the differences of SI between measured and calculated high b-value images and the lesion contrast (SI lesion/liver) were computed. ResultsWith increasing upper measured b-values, the CV in cDWI and vcDWI decreased (CV in the liver in cb1500: 0.42 with b100/600 s/mm2 and 0.28 with b100/b1000s/mm2) while the differences of measured and calculated b-value images decreased (in the liver in cb1500: 30.7% with b = 100/600 s/mm2, 19.7% with b100/b1000s/mm2). In diffusion-restricted lesions, lesion contrast was at least 1.6 in cb1000 and 1.4 in cb1500, respectively, with an upper measured b-value of b = 800 s/mm2 and 2.1 for vcDWI with an upper measured b-value of b = 1000s/mm2. Overall, the lesion contrast was superior in cb1500 and vcDWI compared to cb1000 (15% and 11%, respectively). ConclusionMeasuring higher upper b-values seems to lead to more precise computed high b-value images and a decrease of CV. vcDWI provides a comparable lesion contrast to b = 1500s/mm2 and offers additionally the reduction of T2 shine-through effects. For vcDWI, measuring b = 1000s/mm2 as upper b-value seems to be necessary to guarantee good lesion visibility in the liver based on our preliminary results.

Experimentally Researched Statistical Characteristics of Aerosol Lidar–Recorded Laser Signal at the Eye-Safe Wavelength of 0.355 μm

To date the developed and newly created laser systems (radar, range finding, altimetry, etc.) use (to ensure eye safety) the infrared (> 1.4 μm) range. At the same time, an ultraviolet range is of interest too. The most promising here is a wavelength of 0.355 μm (third harmonic of a YAG laser: Nd), since the atmospheric ozone absorption is still low (in the spectral range of 0.25 - 0.38 μm ozone absorption is increased with decreasing wavelength). The paper presents an experimental study of the statistical characteristics of lidar signals at the eye-safe wavelength of 0.355μm. As the statistical characteristics, were used a mean value and a standard deviation of lidar signal. Aerosol inhomogeneity available in the atmosphere has impact on the statistical characteristics of the received signal. Thus, the analysis of these characteristics allows us to view that in the atmosphere there is aerosol inhomogeneity. The article shows that the standard deviation of the signal caused by inhomogeneity is much higher than the standard deviation caused by the shot noise of signal and background as well as by the noise of electronic circuit. The article gives examples of the field of a coefficient of variations. After filtration the field also permits viewing the available atmospheric inhomogeneity of the backscattered signal.

Quantified coronary frequency domain optical coherence tomography signal analysis for the evaluation of erythrocyte-rich thrombus: ex-vivo validation study.

Previous study has demonstrated that erythrocyte-rich thrombi contain more inflammatory cells and reflect high thrombus burden, leading to impaired myocardial reperfusion in myocardial infarction. The aim of this study is to investigate the utility of quantified frequency domain optical coherence tomography (FD-OCT) signal analysis in evaluating the erythrocyte-rich thrombus with ex-vivo materials. We evaluated 54 specimens of coronary artery thrombus obtained by thrombectomy catheter from 8 patients who underwent primary percutaneous coronary intervention. The thrombi were immersed in saline immediately after thrombectomy and FD-OCT image acquisition was performed ex-vivo. Quantitative analysis for all contiguous frames was performed by the dedicated automated software (OCT system software, Light Lab Inc.). For the maximum thrombus area, mean signal intensity (MSI) and normalized standard deviation of signal (NSD) was evaluated. All thrombi were stained using double staining of phosphotungstic acid-hematoxylin and eosin to enable automatic extraction of erythrocyte from fibrin. Computer-assisted analysis was performed using dedicated image processing software (WinROOF, Mitani Corp., Tokyo, Japan) for color identification of the erythrocyte area. Erythrocyte-rich thrombus, defined as % erythrocyte [(erythrocyte area/total thrombus area) × 100] ≥ 10%, showed significantly lower MSI [4.39 ± 0.24 vs. 4.74 ± 0.35, p = 0.002] than that of <10%. The cut-off point for prediction of erythrocyte-rich thrombus was defined as MSI ≤ 4.56, sensitivity: 87.5%, specificity: 82.9%, area under the curve: 0.836, respectively). The present ex-vivo study suggested the utility of quantified FD-OCT signal analysis on the detection of erythrocyte-rich thrombus.

Particle cluster sizing in downer units. Applicable methodology across downer scale units

Data analysis from downers requires a comprehensive methodology in order to set the data baseline. This analysis can be accomplished by using a mathematical approach based on solid mass balances and an iterative calculation. Each signal baseline is defined asX+n∙σx, where X is the signal average; σx is the signal standard deviation; and “n” is a baseline parameter. By using this methodology, the noise resulting from secondary reflections is eliminated and all valuable data is kept in the time series.More than 500 experiments with >50 million data records were obtained using two independent gas-solid circulating fluidized bed downer units of 3m height and two different diameters (2.54-cm and 5.08-cm ID). The solid used was a FCC catalyst with a mean particle diameter of 84.42μm and a particle density of 1722kg/m3. Measurements were effected using a CREC-GS-Optiprobe, an optical sensor equipped with a GRIN lens. This lens focus radiation in a 118±34μm diameter highly irradiated volume, placed at 4.5mm away from the sensor tip. CREC-GS-Optiprobe sensors do not require calibration and offer minimum intrusion.On the basis of the data obtained, this study reports a valuable methodology, applicable to downer units of different diameters. This leads for the two downers studied to particle clusters with very close asymmetric distributions.

Examining Sample Rate, Sample Time, and Test Replication for Reducing Uncertainty in Steady Timewise Experiments

This paper presents the ways that sample rate, sample time, and number of test replications can affect the random uncertainty in a measurement. Typical steady timewise experiments seek the average values of measured variables. Even in this case, sample rate and sample time can affect the signal standard deviations and yield different random uncertainty estimates. In addition, many random error sources vary slowly relative to the test time and take on a single value. Test replications can convert systematic uncertainties to random uncertainties by allowing their values to change from test to test. The goal is to record individual tests at a sample rate and time that capture the short timescale error sources, and to replicate tests on the scale of long timescale error sources. This paper presents how to leverage these effects to reduce the overall uncertainty of a measured result without increasing the cost of the experiment.

Application of a self-developed proton transfer reaction-mass spectrometer to on-line monitoring trace volatile organic compounds in ambient air

Real-time and on-line monitoring volatile organic compounds(VOCs) are valuable for real-time evaluating air quality and monitoring the key source of pollution. A self-developed proton transfer reaction-mass spectrometer( PTR-MS) was constructed and applied to on-line monitoring trace VOCs in ambient air in Hefei. With the help of a self-developed catalytic converter, the background signal of the instrument was detected and the stability of the instrument was evaluated. The relative standard deviation of signal at m/z 21 was only 0.74% and the detection limit of PTR-MS was 97 part per trillion(97×10–12, volume ratio). As a case of the air monitoring in Hefei, the ambient air at Dongpu reservoir spot was on-line monitored for 13 d with our self-developed PTR-MS. Meanwhile, a solid-phase micro-extraction(SPME) technique coupled to gas chromatography-mass spectrometry/mass spectrometry (GC-MS/MS) was also used for the off-line detection of the air. The results show that our self-developed PTR-MS can be used for the on-line and long-term monitoring of VOCs in air at part per trillion level, and the change trend of VOCs concentration monitored with PTR-MS was consistent with that detected with the conventional SPME-GC-MS. This self-developed PTR-MS can fully satisfy the requirements of air quality monitoring and real-time monitoring of the key pollution sources.

Hardware Efficient Automatic Thresholding for NEO-Based Neural Spike Detection.

The nonlinear energy operator (NEO) algorithm has been commonly implemented in hardware for neural spike detection. However, the traditional method to set the threshold is sensitive to the spike firing rate. In this paper, a new approach is presented to automatically set the threshold, in real time, in a manner that is robust to the spike firing rate and suitable for a neural implant. The presented threshold calculation method statistically analyzes the neural signal standard deviation and root-mean-square frequency and can update the threshold of each channel sequentially every few seconds. Hardware efficient architectures to estimate the threshold calculation statistical parameters are also presented. This automatic thresholding method for NEO spike detection shows robust performance for firing rates from 10 to 100, occupies only 0.021mm2 in 130nm CMOS, and consumes only 50 nW in simulations with a 20-kHz clock.

Development of a frequency regulation duty-cycle for standardized energy storage performance testing

The US DOE Protocol for uniformly measuring and expressing the performance of energy storage systems, first developed in 2012 through inclusive working group activities, provides standardized methodologies for evaluating an energy storage system's ability to supply specific services to electrical grids. This article elaborates on the data and decisions behind the duty-cycle used for frequency regulation in this protocol. Analysis of a year of publicly available frequency regulation control signal data from a utility was considered in developing the representative signal for this use case. This showed that signal standard deviation can be used as a metric for aggressiveness or rigor. From these data, we select representative 2h long signals that exhibit nearly all of dynamics of actual usage under two distinct regimens, one for average use and the other for highly aggressive use. These results were combined into a 24-h duty-cycle comprised of average and aggressive segments. The benefits and drawbacks of the selected duty-cycle are discussed along with its potential implications to the energy storage industry.

Hilbert Transform based Fuzzy Expert System for Diagnosing and Classifying Power Quality Disturbances

This paper presents a new technique for diagnosis and classification of power quality disturbances. The proposed method applies Hilbert transform to analyze the distorted voltage waveforms and then extract their features. The distorted voltage waveforms are generated by Matlab simulink on the test system. The extracted input features such as standard deviation and variances are given as inputs to the fuzzy-expert system that uses some rules to classify the Power Quality disturbances. Fuzzy classifier has been constructed to classify both the single and combined form power quality disturbances. The results clearly show that the proposed method has the ability to diagnosize and classify Power Quality problems. The results obtained by the proposed method are validated by comparing them against Hilbert Transform based neural classifiers. General Terms – Real signal Hilbert transform signal Shifting operator Shifting the negative frequency of Envelop signal of – Instantaneous phase signal of Envelop mean value of the signal Variances of the envelop signal Standard deviation of the signal

Optimization of liquid jet system for laser-induced breakdown spectroscopy analysis

A complex optimization of geometrical and temporal parameters of a jet system (developed in Laser-induced breakdown spectroscopy (LIBS) laboratory of Brno University of Technology) for direct elemental analysis of samples in a liquid state of matter using LIBS was carried out. First, the peristaltic pump was synchronized with the flashlamp of the ablation laser, which reduced variation of the ablated sample amount. Also, the fluctuation of the laser ray angle incident on the jet surface was diminished. Such synchronization reduced signal standard deviations and thus increased repeatability of the measurements. Then, laser energy and distance of the focusing lens from the sample were optimized. The gate delay time and the gate width were optimized for single pulse (SP) experiments; the gate delay time and the inter-pulse delay were optimized for the use of double pulse (DP) variant. Results were assessed according to the highest signal to noise ratios and the lowest relative standard deviations of the signal. The sensitivity of the single pulse and the double pulse LIBS for the detection of heavy metals traces, copper (Cu i at 324.754 nm) and lead (Pb i at 405.781 nm), in aqueous solution of copper (ii) sulfate and lead (ii) acetate, was estimated in terms of limits of detection (LODs). As a result, sensitivity improvement of DP LIBS system was observed, the LOD of Cu obtained with DP was calculated 40% lower than LOD gained from SP technique.

Global signal modulation of single-trial fMRI response variability: Effect on positive vs negative BOLD response relationship

In functional magnetic resonance imaging (fMRI), the relationship between positive BOLD responses (PBRs) and negative BOLD responses (NBRs) to stimulation is potentially informative about the balance of excitatory and inhibitory brain responses in sensory cortex. In this study, we performed three separate experiments delivering visual, motor or somatosensory stimulation unilaterally, to one side of the sensory field, to induce PBR and NBR in opposite brain hemispheres. We then assessed the relationship between the evoked amplitudes of contralateral PBR and ipsilateral NBR at the level of both single-trial and average responses. We measure single-trial PBR and NBR peak amplitudes from individual time-courses, and show that they were positively correlated in all experiments. In contrast, in the average response across trials the absolute magnitudes of both PBR and NBR increased with increasing stimulus intensity, resulting in a negative correlation between mean response amplitudes. Subsequent analysis showed that the amplitude of single-trial PBR was positively correlated with the BOLD response across all grey-matter voxels and was not specifically related to the ipsilateral sensory cortical response. We demonstrate that the global component of this single-trial response modulation could be fully explained by voxel-wise vascular reactivity, the BOLD signal standard deviation measured in a separate resting-state scan (resting state fluctuation amplitude, RSFA). However, bilateral positive correlation between PBR and NBR regions remained. We further report that modulations in the global brain fMRI signal cannot fully account for this positive PBR–NBR coupling and conclude that the local sensory network response reflects a combination of superimposed vascular and neuronal signals. More detailed quantification of physiological and noise contributions to the BOLD signal is required to fully understand the trial-by-trial PBR and NBR relationship compared with that of average responses.

FPGA implementation of a high‐speed, real‐time, windowed standard deviation filter

Characterisation of the standard deviation of a time-series signal has uncommon, yet widespread applications. The usual requirement for a representation of signal standard deviation in real-time implies a high computation speed. A method based on a field programmable gate array (FPGA) implementation is presented. The technique is benchmarked against conventional computational approaches and shows a single windowed standard deviation update calculation of a 16 bit sample can be achieved in 11 ns on a modern CPU. The FPGA implementation is found to be superior to all other approaches examined with an operation time of below 10 ns, and thus provides a useful tool for the real-time measurement of the standard deviation of signals above 100 MHz.

Railway Ballast Fouling Detection Using GPR Data: Introducing a Combined Time–Frequency and Discrete Wavelet Techniques

ABSTRACTGround‐penetrating radar has been recently used for quantifying railway ballast fouling conditions. This paper discusses two approaches for processing the railway ballast ground‐penetrating radar signal: the discrete wavelet transform method and the short‐time Fourier transform method. Ground‐penetrating radar field data collected in Wyoming in 2007 by two 2‐GHz air‐coupled antennas were used to verify both approaches. The collected ground‐penetrating radar signals were processed by discrete wavelet transform. The signal standard deviation values were fitted by linear regression of the fouling indices, which were calculated using ground‐truth sieving data. The spectrogram generated by short‐time Fourier transform was used to analyse the ballast fouling condition level by comparing the energy attenuation speed. It was observed that both discrete wavelet transform and short‐time Fourier transform approaches can be used to analyse ballast fouling conditions. The short‐time Fourier transform method is easier to visualize, whereas the discrete wavelet transform approach can more quantitatively predict the fouling level. However, the discrete wavelet transform method provides the fouling condition of the whole ballast layer, whereas the spectrogram provides the fouling condition with respect to the depth profile.

Key factors dominating the neural coding preference to 1/f signal

Experiments have demonstrated that cortical and sensory neurons prefer to response to signals with characteristics of long-term correlation or 1/f noise feature better than signals with no correlation like white-noise-type [1]. In order to study the underlying mechanism, we built up a cortical neuronal model [2] based on Hodgkin-Huxley theory to study the correlations between neuron kinetics and signal statistics. Interestingly, we observed that (see Figure 1a,b) white-noise-type signal (cutoff frequency >10000Hz) is hard to induce action potentials unless signals with very strong intensity while 1/f signal and low-pass filtered white noise type signal (cutoff frequency <1000Hz) can easily induce action potentials with high firing rate at low signal intensity (quantified by signal standard deviation STD), see Figure ​Figure1a.1a. Moreover, the half-height duration of action potentials is also varying with more sensitivity to 1/f signal or filtered white noise than pure white noise (Figure ​(Figure1b).1b). In addition, neural firing rate and spike duration are more sensitive to 1/f signal than filtered white noise. This is a clear positive evidence of neuronal coding preference to signals with long-term correlations. Figure 1 Interactions among neuron kinetics and signal statistics. a) Firing rate vs signal STD for white noise, filtered white noise and 1/f noise. b) Half-height duration of action potentials vs signal STD for white noise, filtered white noise and 1/f noise. ... To further reveal the key factors dominating the preference of neuronal dynamics to colored and white noise, we systematically varied the values of Na+ and K+ channel time constants and channel rate constants. Figure ​Figure1c1c showed that the firing rate doesn't change much as a function of sodium activation time constant (τm) for a given 1/f signal (STD = 1). However, the firing rate decrease dramatically with an increase of sodium inactivation time constant (τh) or potassium activation time constant (τn) for this signal. For very large τh and τn, neuronal response dynamics start to be saturate for all signals, see Figure ​Figure1d.1d. These results indicate that both sodium inactivation time constant (τh) and potassium activation time constant (τn) may be the key factors dominating neural coding preference to signals with different correlated statistic features. In sum, this study demonstrated that ionic channel time constant of sodium inactivation and potassium activation may be the dominating factors accounting for the neuronal spiking sensitivity favorable to signals with temporal correlations.

Parametric estimation of sample entropy for physical activity recognition.

Insufficient amount of physical activity, and hence storage of calories may lead depression, obesity, cardiovascular diseases, and diabetes. The amount of consumed calorie depends on the type of activity. The recognition of physical activity is very important to estimate the amount of calories spent by a subject every day. There are some research works already published in the literature for activity recognition through accelerometers (body worn sensors). The accuracy of any recognition system depends on the robustness of selected features and classifiers. The typical features reported for most physical activities recognitions are autoregressive coefficients (ARcoeffs), signal magnitude area (SMA), tilt angle (TA), and standard deviation (STD). In this study, we have studied the feasibility of using single value of sample entropy estimated parametrically (SETH) of an AR model instead of ARcoeffs. After feasibility study, we also compared the recognition accuracies between two popular classifiers ı.e. artificial neural network (ANN) and support vector machines (SVM). The recognition accuracies using linear structure (where all types of activities are classified using a single classifier) and hierarchical structure (where activities are first divided into static and dynamic events, and then activities of each event are classified in the second stage). The study showed that the use of SETH provides similar recognition accuracy (69.82%) as provided by ARcoeffs (67.67%) using ANN. The linear structure of SVM performs better (average accuracy of SVM: 98.22%) than linear ANN (average accuracy with ANN: 94.78%). The use of hierarchical structure of ANN increases the average recognition accuracy of static activities to about 100%. However, no significant changes are observed using hierarchical SVM than the linear one.

Rugged, Scramjet Inlet Temperature and Velocity Sensor: Design and Ground Test

The design and testing of a gas temperature and velocity scramjet engine inlet sensor based on tunable diode laser absorption spectroscopy of oxygen is discussed. The robustness and stability of the optical design necessitated by the dynamic vibrational and thermal environment were balanced against the need for compact size and very high signal-to-noise ratio in a low-pressure, high-altitude flight regime. Test results demonstrated very good mechanical and thermal stability and an adequate signal-to-noise ratio for altitudes up to 30 km with predicted absorbance peaks greater than the signal standard deviation by a factor of more than 3.