ABSTRACT
Introduction: Inertial Measurement Units (IMU) require a sensor-to-segment calibration procedure in order to compute anatomically accurate joint angles and, thereby, be employed in healthcare and rehabilitation. Research literature proposes several algorithms to address this issue. However, determining an optimal calibration procedure is challenging due to the large number of variables that affect elbow joint angle accuracy, including 3D joint axis, movement performed, complex anatomy, and notable skin artefacts. Therefore, this paper aims to compare three types of calibration techniques against an optical motion capture reference system during several movement tasks to provide recommendations on the most suitable calibration for the elbow joint. Methods: Thirteen healthy subjects were instrumented with IMU sensors and optical marker clusters. Each participant performed a series of static poses and movements to calibrate the instruments and, subsequently, performed single-plane and multi-joint tasks. The metrics used to evaluate joint angle accuracy are Range of Motion (ROM) error, Root Mean Squared Error (RMSE), and offset. We performed a three-way RM ANOVA to evaluate the effect of joint axis and movement task on three calibration techniques: N-Pose (NP), Functional Calibration (FC) and Manual Alignment (MA). Results: Despite small effect sizes in ROM Error, NP displayed the least precision among calibrations due to interquartile ranges as large as 24.6°. RMSE showed significant differences among calibrations and a large effect size where MA performed best (RMSE = 6.3°) and was comparable with FC (RMSE = 7.2°). Offset showed a large effect size in the calibration*axes interaction where FC and MA performed similarly. Conclusion: Therefore, we recommend MA as the preferred calibration method for the elbow joint due to its simplicity and ease of use. Alternatively, FC can be a valid option when the wearer is unable to hold a predetermined posture.
ABSTRACT
Human skin classification is an essential task for several machine vision applications such as human-machine interfaces, people/object tracking, and classification. In this paper, we describe a hybrid CMOS/memristor vision sensor architecture embedding skin detection over a wide dynamic range. In-sensor RGB to r g-chromaticity color-space conversion is executed on-the-fly through a pixel-level automatic exposure time control. Each pixel of the array delivers two pre-filtered analog signals, the r and g values, suitable for being efficiently classified as skin or non-skin through an analog memristive neural network (NN), without the need for any further signal processing. Moreover, we study the NN performance and theorize how it should be added in the hardware. The skin classifier is organized in an array of column-level memristor-based NN to exploit the nano-scale device characteristics and non-volatile analog memory capabilities, making the proposed sensor architecture highly flexible, customizable for various use-case scenarios, and low-power. The output is a skin bitmap that is robust against variations of the illuminant color and intensity.
Subject(s)
Neural Networks, Computer , Humans , Skin , Color , Equipment DesignABSTRACT
Recent studies show that the integrity of core perceptual and cognitive functions may be tested in a short time with Steady-State Visual Evoked Potentials (SSVEP) with low stimulation frequencies, between 1 and 10 Hz. Wearable EEG systems provide unique opportunities to test these brain functions on diverse populations in out-of-the-lab conditions. However, they also pose significant challenges as the number of EEG channels is typically limited, and the recording conditions might induce high noise levels, particularly for low frequencies. Here we tested the performance of Normalized Canonical Correlation Analysis (NCCA), a frequency-normalized version of CCA, to quantify SSVEP from wearable EEG data with stimulation frequencies ranging from 1 to 10 Hz. We validated NCCA on data collected with an 8-channel wearable wireless EEG system based on BioWolf, a compact, ultra-light, ultra-low-power recording platform. The results show that NCCA correctly and rapidly detects SSVEP at the stimulation frequency within a few cycles of stimulation, even at the lowest frequency (4 s recordings are sufficient for a stimulation frequency of 1 Hz), outperforming a state-of-the-art normalized power spectral measure. Importantly, no preliminary artifact correction or channel selection was required. Potential applications of these results to research and clinical studies are discussed.
Subject(s)
Brain-Computer Interfaces , Wearable Electronic Devices , Electroencephalography/methods , Evoked Potentials, Visual , Canonical Correlation Analysis , Photic Stimulation/methods , AlgorithmsABSTRACT
Electroencephalogram (EEG) data are typically affected by artifacts. The detection and removal of bad channels (i.e., with poor signal-to-noise ratio) is a crucial initial step. EEG data acquired from different populations require different cleaning strategies due to the inherent differences in the data quality, the artifacts' nature, and the employed experimental paradigm. To deal with such differences, we propose a robust EEG bad channel detection method based on the Local Outlier Factor (LOF) algorithm. Unlike most existing bad channel detection algorithms that look for the global distribution of channels, LOF identifies bad channels relative to the local cluster of channels, which makes it adaptable to any kind of EEG. To test the performance and versatility of the proposed algorithm, we validated it on EEG acquired from three populations (newborns, infants, and adults) and using two experimental paradigms (event-related and frequency-tagging). We found that LOF can be applied to all kinds of EEG data after calibrating its main hyperparameter: the LOF threshold. We benchmarked the performance of our approach with the existing state-of-the-art (SoA) bad channel detection methods. We found that LOF outperforms all of them by improving the F1 Score, our chosen performance metric, by about 40% for newborns and infants and 87.5% for adults.
Subject(s)
Electroencephalography , Signal Processing, Computer-Assisted , Adult , Algorithms , Artifacts , Electroencephalography/methods , Humans , Infant, Newborn , Signal-To-Noise RatioABSTRACT
Policy makers have implemented multiple non-pharmaceutical strategies to mitigate the COVID-19 worldwide crisis. Interventions had the aim of reducing close proximity interactions, which drive the spread of the disease. A deeper knowledge of human physical interactions has revealed necessary, especially in all settings involving children, whose education and gathering activities should be preserved. Despite their relevance, almost no data are available on close proximity contacts among children in schools or other educational settings during the pandemic. Contact data are usually gathered via Bluetooth, which nonetheless offers a low temporal and spatial resolution. Recently, ultra-wideband (UWB) radios emerged as a more accurate alternative that nonetheless exhibits a significantly higher energy consumption, limiting in-field studies. In this paper, we leverage a novel approach, embodied by the Janus system that combines these radios by exploiting their complementary benefits. The very accurate proximity data gathered in-field by Janus, once augmented with several metadata, unlocks unprecedented levels of information, enabling the development of novel multi-level risk analyses. By means of this technology, we have collected real contact data of children and educators in three summer camps during summer 2020 in the province of Trento, Italy. The wide variety of performed daily activities induced multiple individual behaviors, allowing a rich investigation of social environments from the contagion risk perspective. We consider risk based on duration and proximity of contacts and classify interactions according to different risk levels. We can then evaluate the summer camps' organization, observe the effect of partition in small groups, or social bubbles, and identify the organized activities that mitigate the riskier behaviors. Overall, we offer an insight into the educator-child and child-child social interactions during the pandemic, thus providing a valuable tool for schools, summer camps, and policy makers to (re)structure educational activities safely.
ABSTRACT
Electroencephalography (EEG) is arising as a valuable method to investigate neurocognitive functions shortly after birth. However, obtaining high-quality EEG data from human newborn recordings is challenging. Compared to adults and older infants, datasets are typically much shorter due to newborns' limited attentional span and much noisier due to non-stereotyped artifacts mainly caused by uncontrollable movements. We propose Newborn EEG Artifact Removal (NEAR), a pipeline for EEG artifact removal designed explicitly for human newborns. NEAR is based on two key steps: 1) A novel bad channel detection tool based on the Local Outlier Factor (LOF), a robust outlier detection algorithm; 2) A parameter calibration procedure for adapting to newborn EEG data the algorithm Artifacts Subspace Reconstruction (ASR), developed for artifact removal in mobile adult EEG. Tests on simulated data showed that NEAR outperforms existing methods in removing representative newborn non-stereotypical artifacts. NEAR was validated on two developmental populations (newborns and 9-month-old infants) recorded with two different experimental designs (frequency-tagging and ERP). Results show that NEAR artifact removal successfully reproduces established EEG responses from noisy datasets, with a higher statistical significance than the one obtained by existing artifact removal methods. The EEGLAB-based NEAR pipeline is freely available at https://github.com/vpKumaravel/NEAR.
Subject(s)
Artifacts , Signal Processing, Computer-Assisted , Adult , Algorithms , Electroencephalography/methods , Humans , Infant , Infant, Newborn , MovementABSTRACT
Light-weight, minimally-obtrusive mobile EEG systems with a small number of electrodes (i.e., low-density) allow for convenient monitoring of the brain activity in out-of-the-lab conditions. However, they pose a higher risk for signal contamination with non-stereotypical artifacts due to hardware limitations and the challenging environment where signals are collected. A promising solution is Artifacts Subspace Reconstruction (ASR), a component-based approach that can automatically remove non-stationary transient-like artifacts in EEG data. Since ASR has only been validated with high-density systems, it is unclear whether it is equally efficient on low-density portable EEG. This paper presents a complete analysis of ASR performance based on clean and contaminated datasets acquired with BioWolf, an Ultra-Low-Power system featuring only eight channels, during SSVEP sessions recorded from six adults. Empirical results show that even with such few channels, ASR efficiently corrects artifacts, enabling an overall enhancement of up to 40% in SSVEP response. Furthermore, by choosing the optimal ASR parameters on a single-subject basis, SSVEP response can be further increased to more than 45%. These results suggest that ASR is a viable and robust method for online automatic artifact correction with low-density BCI systems in real-life scenarios.
Subject(s)
Artifacts , Wearable Electronic Devices , Algorithms , Electroencephalography , Signal Processing, Computer-AssistedABSTRACT
BACKGROUND: Emerging sensing and communication technologies are contributing to the development of many motor rehabilitation programs outside the standard healthcare facilities. Nowadays, motor rehabilitation exercises can be easily performed and monitored even at home by a variety of motion-tracking systems. These are cheap, reliable, easy-to-use, and allow also remote configuration and control of the rehabilitation programs. The two most promising technologies for home-based motor rehabilitation programs are inertial wearable sensors and video-based motion capture systems. METHODS: In this paper, after a thorough review of the relevant literature, an original experimental analysis is reported for two corresponding commercially available solutions, a wearable inertial measurement unit and the Kinect, respectively. For the former, a number of different algorithms for rigid body pose estimation from sensor data were also tested. Both systems were compared with the measurements obtained with state-of-the-art marker-based stereophotogrammetric motion analysis, taken as a gold-standard, and also evaluated outside the lab in a home environment. RESULTS: The results in the laboratory setting showed similarly good performance for the elementary large motion exercises, with both systems having errors in the 3-8 degree range. Usability and other possible limitations were also assessed during utilization at home, which revealed additional advantages and drawbacks for the two systems. CONCLUSIONS: The two evaluated systems use different technology and algorithms, but have similar performance in terms of human motion tracking. Therefore, both can be adopted for monitoring home-based rehabilitation programs, taking adequate precautions however for operation, user instructions and interpretation of the results.
Subject(s)
Exercise Therapy , Mechanical Phenomena , Monitoring, Physiologic/instrumentation , Motor Activity/physiology , Wearable Electronic Devices , Biomechanical Phenomena , HumansABSTRACT
Recent research showed that visual cueing can have both beneficial and detrimental effects on handwriting of patients with Parkinson's disease (PD) and healthy controls depending on the circumstances. Hence, using other sensory modalities to deliver cueing or feedback may be a valuable alternative. Therefore, the current study compared the effects of short-term training with either continuous visual cues or intermittent intelligent verbal feedback. Ten PD patients and nine healthy controls were randomly assigned to one of these training modes. To assess transfer of learning, writing performance was assessed in the absence of cueing and feedback on both trained and untrained writing sequences. The feedback pen and a touch-sensitive writing tablet were used for testing. Both training types resulted in improved writing amplitudes for the trained and untrained sequences. In conclusion, these results suggest that the feedback pen is a valuable tool to implement writing training in a tailor-made fashion for people with PD. Future studies should include larger sample sizes and different subgroups of PD for long-term training with the feedback pen.
ABSTRACT
EEG is a standard non-invasive technique used in neural disease diagnostics and neurosciences. Frequency-tagging is an increasingly popular experimental paradigm that efficiently tests brain function by measuring EEG responses to periodic stimulation. Recently, frequency-tagging paradigms have proven successful with low stimulation frequencies (0.5-6Hz), but the EEG signal is intrinsically noisy in this frequency range, requiring heavy signal processing and significant human intervention for response estimation. This limits the possibility to process the EEG on resource-constrained systems and to design smart EEG based devices for automated diagnostic. We propose an algorithm for artifact removal and automated detection of frequency tagging responses in a wide range of stimulation frequencies, which we test on a visual stimulation protocol. The algorithm is rooted on machine learning based pattern recognition techniques and it is tailored for a new generation parallel ultra low power processing platform (PULP), reaching performance of more that 90% accuracy in the frequency detection even for very low stimulation frequencies (<1Hz) with a power budget of 56mW.
Subject(s)
Electroencephalography/methods , Machine Learning , Photic Stimulation/methods , Algorithms , Artifacts , HumansABSTRACT
Poliarticulated prosthetic hands represent a powerful tool to restore functionality and improve quality of life for upper limb amputees. Such devices offer, on the same wearable node, sensing and actuation capabilities, which are not equally supported by natural interaction and control strategies. The control in state-of-the-art solutions is still performed mainly through complex encoding of gestures in bursts of contractions of the residual forearm muscles, resulting in a non-intuitive Human-Machine Interface (HMI). Recent research efforts explore the use of myoelectric gesture recognition for innovative interaction solutions, however there persists a considerable gap between research evaluation and implementation into successful complete systems. In this paper, we present the design of a wearable prosthetic hand controller, based on intuitive gesture recognition and a custom control strategy. The wearable node directly actuates a poliarticulated hand and wirelessly interacts with a personal gateway (i.e., a smartphone) for the training and personalization of the recognition algorithm. Through the whole system development, we address the challenge of integrating an efficient embedded gesture classifier with a control strategy tailored for an intuitive interaction between the user and the prosthesis. We demonstrate that this combined approach outperforms systems based on mere pattern recognition, since they target the accuracy of a classification algorithm rather than the control of a gesture. The system was fully implemented, tested on healthy and amputee subjects and compared against benchmark repositories. The proposed approach achieves an error rate of 1.6% in the end-to-end real time control of commonly used hand gestures, while complying with the power and performance budget of a low-cost microcontroller.
Subject(s)
Gestures , Algorithms , Amputees , Artificial Limbs , Electromyography , Hand , Humans , Pattern Recognition, Automated , Prostheses and Implants , Quality of LifeABSTRACT
Embedded vision systems are smart energy-efficient devices that capture and process a visual signal in order to extract high-level information about the surrounding observed world. Thanks to these capabilities, embedded vision systems attract more and more interest from research and industry. In this work, we present a novel low-power optical embedded system tailored to detect the human skin under various illuminant conditions. We employ the presented sensor as a smart switch to activate one or more appliances connected to it. The system is composed of an always-on low-power RGB color sensor, a proximity sensor, and an energy-efficient microcontroller (MCU). The architecture of the color sensor allows a hardware preprocessing of the RGB signal, which is converted into the rg space directly on chip reducing the power consumption. The rg signal is delivered to the MCU, where it is classified as skin or non-skin. Each time the signal is classified as skin, the proximity sensor is activated to check the distance of the detected object. If it appears to be in the desired proximity range, the system detects the interaction and switches on/off the connected appliances. The experimental validation of the proposed system on a prototype shows that processing both distance and color remarkably improves the performance of the two separated components. This makes the system a promising tool for energy-efficient, touchless control of machines.
Subject(s)
Colorimetry/instrumentation , Man-Machine Systems , Micro-Electrical-Mechanical Systems/instrumentation , Pattern Recognition, Automated/methods , Signal Processing, Computer-Assisted/instrumentation , Skin/anatomy & histology , Electric Power Supplies , Equipment Design , Equipment Failure Analysis , Reproducibility of Results , Sensitivity and Specificity , TransducersABSTRACT
Wearable devices offer interesting features, such as low cost and user friendliness, but their use for medical applications is an open research topic, given the limited hardware resources they provide. In this paper, we present an embedded solution for real-time EMG-based hand gesture recognition. The work focuses on the multi-level design of the system, integrating the hardware and software components to develop a wearable device capable of acquiring and processing EMG signals for real-time gesture recognition. The system combines the accuracy of a custom analog front end with the flexibility of a low power and high performance microcontroller for on-board processing. Our system achieves the same accuracy of high-end and more expensive active EMG sensors used in applications with strict requirements on signal quality. At the same time, due to its flexible configuration, it can be compared to the few wearable platforms designed for EMG gesture recognition available on market. We demonstrate that we reach similar or better performance while embedding the gesture recognition on board, with the benefit of cost reduction. To validate this approach, we collected a dataset of 7 gestures from 4 users, which were used to evaluate the impact of the number of EMG channels, the number of recognized gestures and the data rate on the recognition accuracy and on the computational demand of the classifier. As a result, we implemented a SVM recognition algorithm capable of real-time performance on the proposed wearable platform, achieving a classification rate of 90%, which is aligned with the state-of-the-art off-line results and a 29.7 mW power consumption, guaranteeing 44 hours of continuous operation with a 400 mAh battery.
Subject(s)
Electromyography/instrumentation , Gestures , Pattern Recognition, Automated/methods , Algorithms , Biomedical Engineering/instrumentation , Clothing , Electromyography/methods , Equipment Design , Forearm/physiology , Humans , Muscle, Skeletal/physiology , Reproducibility of ResultsABSTRACT
A key design challenge for successful wireless sensor network (WSN) deployment is a good balance between the collected data resolution and the overall energy consumption. In this paper, we present a WSN solution developed to efficiently satisfy the requirements for long-term monitoring of a historical building. The hardware of the sensor nodes and the network deployment are described and used to collect the data. To improve the network's energy efficiency, we developed and compared two approaches, sharing similar sub-sampling strategies and data reconstruction assumptions: one is based on compressive sensing (CS) and the second is a custom data-driven latent variable-based statistical model (LV). Both approaches take advantage of the multivariate nature of the data collected by a heterogeneous sensor network and reduce the sampling frequency at sub-Nyquist levels. Our comparative analysis highlights the advantages and limitations: signal reconstruction performance is assessed jointly with network-level energy reduction. The performed experiments include detailed performance and energy measurements on the deployed network and explore how the different parameters can affect the overall data accuracy and the energy consumption. The results show how the CS approach achieves better reconstruction accuracy and overall efficiency, with the exception of cases with really aggressive sub-sampling policies.
ABSTRACT
In this paper, a system for gait training and rehabilitation for Parkinson's disease (PD) patients in a daily life setting is presented. It is based on a wearable architecture aimed at the provision of real-time auditory feedback. Recent studies have, in fact, shown that PD patients can receive benefit from a motor therapy based on auditory cueing and feedback, as happens in traditional rehabilitation contexts with verbal instructions given by clinical operators. To this extent, a system based on a wireless body sensor network and a smartphone has been developed. The system enables real-time extraction of gait spatio-temporal features and their comparison with a patient's reference walking parameters captured in the lab under clinical operator supervision. Feedback is returned to the user in form of vocal messages, encouraging the user to keep her/his walking behavior or to correct it. This paper describes the overall concept, the proposed usage scenario and the parameters estimated for the gait analysis. It also presents, in detail, the hardware-software architecture of the system and the evaluation of system reliability by testing it on a few subjects.
Subject(s)
Gait , Parkinson Disease/physiopathology , Parkinson Disease/rehabilitation , Telemetry/instrumentation , Algorithms , Calibration , Computer Systems , Feedback , Humans , Male , ThermodynamicsABSTRACT
The control of postural sway depends on the dynamic integration of multi-sensory information in the central nervous system. Augmentation of sensory information, such as during auditory biofeedback (ABF) of the trunk acceleration, has been shown to improve postural control. By means of quantitative electroencephalography (EEG), we examined the basic processes in the brain that are involved in the perception and cognition of auditory signals used for ABF. ABF and Fake ABF (FAKE) auditory stimulations were delivered to 10 healthy naive participants during quiet standing postural tasks, with eyes-open and closed. Trunk acceleration and 19-channels EEG were recorded at the same time. Advanced, state-of-the-art EEG analysis and modeling methods were employed to assess the possibly differential, functional activation, and localization of EEG spectral features (power in α, ß, and γ bands) between the FAKE and the ABF conditions, for both the eyes-open and the eyes-closed tasks. Participants gained advantage by ABF in reducing their postural sway, as measured by a reduction of the root mean square of trunk acceleration during the ABF compared to the FAKE condition. Population-wise localization analysis performed on the comparison FAKE - ABF revealed: (i) a significant decrease of α power in the right inferior parietal cortex for the eyes-open task; (ii) a significant increase of γ power in left temporo-parietal areas for the eyes-closed task; (iii) a significant increase of γ power in the left temporo-occipital areas in the eyes-open task. EEG outcomes supported the idea that ABF for postural control heavily modulates (increases) the cortical activation in healthy participants. The sites showing the higher ABF-related modulation are among the known cortical areas associated with multi-sensory, perceptual integration, and sensorimotor integration, showing a differential activation between the eyes-open and eyes-closed conditions.