Efficacy and safety for combination of t-EMG with O-arm assisted pedicle screw placement in neurofibromatosis type I scoliosis surgery

Background Due to the characteristics of neurofibromatosis type I (NF-1) scoliosis, the precise placement of pedicle screws still remains to be a challenge. Triggered screw electromyography (t-EMG) has been proved to exhibit high sensitivity to identify mal-positioned pedicle screws, but no previous study assessed the combination of t-EMG with O-arm-assisted pedicle screw placement in NF-1 scoliosis surgery. Objective To evaluate efficacy and safety for combination of t-EMG with O-arm-assisted pedicle screw placement in NF-1 scoliosis surgery. Materials and methods From March 2018 to April 2020, sixty-five NF-1 scoliosis patients underwent t-EMG and O-arm-assisted pedicle screw fixation were retrospectively reviewed. The channel classification system was applied to classify the pedicle morphology based on pedicle width measurement by preoperative computed tomography scans. The minimal t-EMG threshold for screw path inspection was used as 8 mA, and operative screw redirection was also recorded. All pedicle screws were verified using a second intraoperative O-arm scan. The correlation between demographic and clinical data with amplitude of t-EMG were also analyzed. Results A total of 652 pedicle screws (T10-S1) in 65 patients were analyzed. The incidence of an absent pedicle (channel classification type C or D morphology) was 150 (23%). Overall, abnormal t-EMG threshold was identified in 26 patients with 48 screws (7.4%), while 16 out of the 48 screws were classified as G0, 14 out of the 48 screws were classified as G1, and 18 out of the 48 screws were classified as G2. The screw redirection rate was 2.8% (18/652). It showed that t-EMG stimulation detected 3 unacceptable mal-positioned screws in 2 patients (G2) which were missed by O-arm scan. No screw-related neurological or vascular complications were observed. Conclusions Combination of t-EMG with O-arm-assisted pedicle screw placement was demonstrated to be a safe and effective method in NF-1 scoliosis surgery. The t-EMG could contribute to detecting the rupture of the medial wall which might be missed by O-arm scan. Combination of t-EMG with O-arm could be recommended for routine use of screw insertion in NF-1 scoliosis surgery.

A Novel Processing Model for P300 Brainwaves Detection

Event-related potential (ERP) is a distinctive pattern of brain activity that is elicited by the brain’s sensitivity and cognition whereas P300 evoked potential changes in cognitive functions. Since P300 wave is a cognitive response across multiple brain channels correlated between the measured electroencephalogram (EEG) and deviant stimulus in a specific period, it requires a suitable signal processing application for interpretation. Moreover, multiple steps of data processing under neuroscience criteria make the P300 reflection difficult to analyze by common methods. Therefore, this study proposes the processing model for brainwave applications based on P300 peak signal detection in multiple brain channels. This study applies 64 channels ERP datasets throughout bandpass filter in fast Fourier transform (FFT) with the specific ranges of signal processing while ERP averaging is applied as a feature extraction method. Furthermore, the experimental metadata is applied with the filtered P300 peak signals in channel classification via a machine learning method, the Decision Tree. The experimental results indicate the accurate mental reflection of P300 evoked potential in different brain channels with high classification accuracy relying on the contrast condition throughout the original data source averaged across the individual electrodes.

Near-channel classifier: symbiotic communication and classification in high-dimensional space

Brain-inspired high-dimensional (HD) computing represents and manipulates data using very long, random vectors with dimensionality in the thousands. This representation provides great robustness for various classification tasks where classifiers operate at low signal-to-noise ratio (SNR) conditions. Similarly, hyperdimensional modulation (HDM) leverages the robustness of complex-valued HD representations to reliably transmit information over a wireless channel, achieving a similar SNR gain compared to state-of-the-art codes. Here, we first propose methods to improve HDM in two ways: (1) reducing the complexity of encoding and decoding operations by generating, manipulating, and transmitting bipolar or integer vectors instead of complex vectors; (2) increasing the SNR gain by 0.2 dB using a new soft-feedback decoder; it can also increase the additive superposition capacity of HD vectors up to 1.7$$\times$$ × in noise-free cases. Secondly, we propose to combine encoding/decoding aspects of communication with classification into a single framework by relying on multifaceted HD representations. This leads to a near-channel classification (NCC) approach that avoids transformations between different representations and the overhead of multiple layers of encoding/decoding, hence reducing latency and complexity of a wireless smart distributed system while providing robustness against noise and interference from other nodes. We provide a use-case for wearable hand gesture recognition with 5 classes from 64 EMG sensors, where the encoded vectors are transmitted to a remote node for either performing NCC, or reconstruction of the encoded data. In NCC mode, the original classification accuracy of 94% is maintained, even in the channel at SNR of 0 dB, by transmitting 10,000-bit vectors. We remove the redundancy by reducing the vector dimensionality to 2048-bit that still exhibits a graceful degradation: less than 6% accuracy loss is occurred in the channel at − 5 dB, and with the interference from 6 nodes that simultaneously transmit their encoded vectors. In the reconstruction mode, it improves the mean-squared error by up to 20 dB, compared to standard decoding, when transmitting 2048-dimensional vectors.

Anchor-Free Localization Using a Deep Neural Network in Wireless Sensor Networks with Multiple Sinks

<div>Wireless sensor networks (WSNs) is one of the vital part of the Internet of Things (IoT) that allow to acquire and provide information from interconnected sensors. Localization-based services are among the most appealing applications associated to the IoT. The deployment of WSNs in the indoor environments and urban areas creates obstacles that lead to the Non-Line-of-Sight (NLOS) propagation. Additionally, the localization accuracy is minimized by the NLOS propagation. The main intention of this paper is to develop an anchor-free node localization approach in multi-sink WSN under NLOS conditions using three key phases such as LOS/NLOS channel classification, range estimation, and anchor-free node localization. The first phase adopts Heuristicbased Deep Neural Network (H-DNN) for LOS/NLOS channel classification. Further, the same H-DNN s used for the range estimation. The hidden neurons of DNN are optimized using the proposed Adaptive Separating Operator-based Elephant Herding Optimization (ASO-EHO) algorithm. The node localization is formulated as a multi-objective optimization problem. The objectives such as localization error, hardware cost, and energy overhead are taken into consideration. ASO-EHO is used for node localization. The suitability of the proposed anchor-free node localization model is validated by comparing over the existing models with diverse counts of nodes. </div>

Anchor-Free Localization Using a Deep Neural Network in Wireless Sensor Networks with Multiple Sinks

<div>Wireless sensor networks (WSNs) is one of the vital part of the Internet of Things (IoT) that allow to acquire and provide information from interconnected sensors. Localization-based services are among the most appealing applications associated to the IoT. The deployment of WSNs in the indoor environments and urban areas creates obstacles that lead to the Non-Line-of-Sight (NLOS) propagation. Additionally, the localization accuracy is minimized by the NLOS propagation. The main intention of this paper is to develop an anchor-free node localization approach in multi-sink WSN under NLOS conditions using three key phases such as LOS/NLOS channel classification, range estimation, and anchor-free node localization. The first phase adopts Heuristicbased Deep Neural Network (H-DNN) for LOS/NLOS channel classification. Further, the same H-DNN s used for the range estimation. The hidden neurons of DNN are optimized using the proposed Adaptive Separating Operator-based Elephant Herding Optimization (ASO-EHO) algorithm. The node localization is formulated as a multi-objective optimization problem. The objectives such as localization error, hardware cost, and energy overhead are taken into consideration. ASO-EHO is used for node localization. The suitability of the proposed anchor-free node localization model is validated by comparing over the existing models with diverse counts of nodes. </div>

Spectrum Occupancy Model Based on Empirical Data for FM Radio Broadcasting in Suburban Environments

It is well-known that the analog FM radio channels in suburban areas are underutilized. Before reallocating the unused channels for other applications, a regulator must analyze the spectrum occupancy. Many researchers proposed the spectrum occupancy models to find vacant spectrum. However, the existing models do not analyze each channel individually. This paper proposes an approach consisting (i) a spectrum measurement strategy, (ii) an appropriate decision threshold, and (iii) criteria for channel classification, to find the unused channels. The measurement strategy monitors each channel’s activity by capturing the power levels of the passband and the guardband separately. The decision threshold is selected depending on the monitored channel’s activity. The criteria classifies the channels based on the passband’s and guardband’s duty cycles. The results show that the proposed channel classification can identify 42 unused channels. If the power levels of wholebands (existing model) were analyzed instead of passband’s and guardband’s duty cycles, only 24 unoccupied channels were found. Furthermore, we propose the interference criteria, based on relative duty cycles across channels, to classify the abnormally used channels into interference sources and interference sinks, which have 16 and 15 channels, respectively. This information helps the dynamic spectrum sharing avoid or mitigate the interferences.

Interpreting reach-scale classifications and the role of spatial-morphological variables in river channel mapping using machine learning algorithms

&lt;p&gt;Over the years, there has been tremendous growth in the literature as regards river channel classifications, however, very few studies have been able to engage the use of remote sensing products in channel classification at the reach-scale level especially by combining reflections from satellite sensors with channel morphological variables. This study aims to identify discriminating spatio-morphological variables using machine learning algorithms and classify site-specific channel types at the reach scale. Each reach was broadly classified based on valley settings (confined, partly confined and unconfined) and channel types (alluvial or bedrock). However, variations and site observations were recorded for site-specific classification purposes. For each reach, Global Positioning System devices were used to geo-locate their endpoints. Standard field instruments were used for cross-sectional measurements and established hydraulic equations for the derived variables. A total of 249 points across 83 reaches were sampled during the fieldwork. Landsat 8 and Sentinel-1 bands were retrieved for days the fieldwork was carried out/for days close to those dates using Google Earth Engine (GEE) platform. Hierarchical cluster analysis, HCA, using Ward&amp;#8217;s linkages was used to provide a classification for the channel types. For the identification of important variables in predicting channel unit types, the random-forest - recursive feature elimination (RF-RFE) algorithm was used using the rfe() function. To identify the best machine learning algorithm, random-forest (rf), support vector machines (svm), multivariate adaptive regression spline (mars) extreme gradient boosting (xgb) and adaptive boosting (adaboost) were used on the training and test data to identify the best performing algorithm. The rfe() feature selection identified five (5) variables that can significantly help in channel unit type identification. The top five variables are dimensionless stream power, slope, width, wetted perimeter and Band 4. Using ROC curve, sensitivity, and specificity, the mars model has the highest ROC curve. Hence, it appears to be the best performing out of the five. However, if the argument is to be based on positive prediction, then any of the models except adaboost will be preferred given their high sensitivity. The HCA using illustrated the clustering structure of the studied reaches by producing five distinct channel classification types distinguished based on width-depth ratio values (high and low). The five distinct channel types are listed as &amp;#160;M1e, M5e, B1, E5b, and E. These codings are based partly on Rosgen&amp;#8217;s classification while, the capital letters (M, B and E) represent mixed channels, bedrock with moderate width-depth ratio and alluvial channels with low width-depth ratio respectively. Numbers 1 and 5 represent bedrocks and sandy beds based on slope variation respectively. The identified channel unit types are a result of the underlying lithology, process-form dynamics and confinement. As streams are expected to respond differently to shocks and recover from damages, it becomes essential to understand these differences in classification which will go a long way in establishing watershed and streamside management guidelines.&lt;/p&gt;

A statistical-based reach scale classification for the lower Tapajós river channel, eastern Amazonia

The Tapajós river is among the largest rivers in the world and has been credited as the main affluent of the lower Amazon River. Geomorphological studies in the Tapajós commonly deal with evolutionary, hydrological, and sedimentological issues. Recently, important advances have been made in understanding the morphology and dynamics in the Tapajós, especially in the confluence zone near Santarém, eastern Amazonian Brazil. However, the lack of an independent channel classification system makes it difficult to integrate data obtained from different sources. This work presents a classification system for the lower Tapajós based on morphometric variables extracted from transversal profiles coupled with radar and optical remote sensing data. We used discriminant analysis of principal components for the first time in fluvial geomorphology to provide a clustering-based geomorphological classification, which is statistically supported. We propose a segmentation of the channel into three distinct sections referred to as narrower channel reach, higher ria reach, and lower ria reach. The results showed that the channel has a distinct morphological pattern in each of these reaches, which can also be observed by the variation along the longitudinal profile. Our findings showed that the most variable hydrological and sedimentological patterns are found in the lower ria reach, while the higher ria reach comprises a canyon-shaped, more stable part of the channel. We discuss the implications of this pattern on the ria’s evolution. The method developed here could be applied to other areas of the Amazon basin, where the lack of data and logistical difficulties carrying out fieldwork are common obstacles to large-scale investigations. The identified reaches are indended to be considered in the sample design of future works and in the formulation of water resources management strategies.