RESUMEN

Human activity recognition (HAR) is pivotal in advancing applications ranging from healthcare monitoring to interactive gaming. Traditional HAR systems, primarily relying on single data sources, face limitations in capturing the full spectrum of human activities. This study introduces a comprehensive approach to HAR by integrating two critical modalities: RGB imaging and advanced pose estimation features. Our methodology leverages the strengths of each modality to overcome the drawbacks of unimodal systems, providing a richer and more accurate representation of activities. We propose a two-stream network that processes skeletal and RGB data in parallel, enhanced by pose estimation techniques for refined feature extraction. The integration of these modalities is facilitated through advanced fusion algorithms, significantly improving recognition accuracy. Extensive experiments conducted on the UTD multimodal human action dataset (UTD MHAD) demonstrate that the proposed approach exceeds the performance of existing state-of-the-art algorithms, yielding improved outcomes. This study not only sets a new benchmark for HAR systems but also highlights the importance of feature engineering in capturing the complexity of human movements and the integration of optimal features. Our findings pave the way for more sophisticated, reliable, and applicable HAR systems in real-world scenarios.

Asunto(s)

RESUMEN

This paper proposes a retinal prosthesis edge detection (RPED) algorithm that can achieve high visual acuity and low power. Retinal prostheses have been used to stimulate retinal tissue by injecting charge via an electrode array, thereby artificially restoring the vision of visually impaired patients. The retinal prosthetic chip, which generates biphasic current pulses, should be located in the foveal area measuring 5 mm × 5 mm. When a high-density stimulation pixel array is realized in a limited area, the distance between the stimulation pixels narrows, resulting in current dispersion and high-power dissipation related to heat generation. Various edge detection methods have been proposed over the past decade to reduce these deleterious effects and achieve high-resolution pixels. However, conventional methods have the disadvantages of high-power consumption and long data processing times because many pixels are activated to detect edges. In this study, we propose a novel RPED algorithm that has a higher visual acuity and less power consumption despite using fewer active pixels than existing techniques. To verify the performance of the devised RPED algorithm, the peak signal-to-noise ratio and structural similarity index map, which evaluates the quantitative numerical value of the image are employed and compared with the Sobel, Canny, and past edge detection algorithms in MATLAB. Finally, we demonstrate the effectiveness of the proposed RPED algorithm using a 1600-pixel retinal stimulation chip fabricated using a 0.35 µm complementary metal-oxide-semiconductor process.

Asunto(s)

RESUMEN

The global prevalence of obesity presents a pressing challenge to public health and healthcare systems, necessitating accurate prediction and understanding for effective prevention and management strategies. This article addresses the need for improved obesity prediction models by conducting a comprehensive analysis of existing machine learning (ML) and deep learning (DL) approaches. This study introduces a novel hybrid model, Attention-based Bi-LSTM (ABi-LSTM), which integrates attention mechanisms with bidirectional Long Short-Term Memory (Bi-LSTM) networks to enhance interpretability and performance in obesity prediction. Our study fills a crucial gap by bridging healthcare and urban planning domains, offering insights into data-driven approaches to promote healthier living within urban environments. The proposed ABi-LSTM model demonstrates exceptional performance, achieving a remarkable accuracy of 96.5% in predicting obesity levels. Comparative analysis showcases its superiority over conventional approaches, with superior precision, recall, and overall classification balance. This study highlights significant advancements in predictive accuracy and positions the ABi-LSTM model as a pioneering solution for accurate obesity prognosis. The implications extend beyond healthcare, offering a precise tool to address the global obesity epidemic and foster sustainable development in smart cities.

RESUMEN

Artificial retinas have revolutionized the lives of many blind people by enabling their ability to perceive vision via an implanted chip. Despite significant advancements, there are some limitations that cannot be ignored. Presenting all objects captured in a scene makes their identification difficult. Addressing this limitation is necessary because the artificial retina can utilize a very limited number of pixels to represent vision information. This problem in a multi-object scenario can be mitigated by enhancing images such that only the major objects are considered to be shown in vision. Although simple techniques like edge detection are used, they fall short in representing identifiable objects in complex scenarios, suggesting the idea of integrating primary object edges. To support this idea, the proposed classification model aims at identifying the primary objects based on a suggested set of selective features. The proposed classification model can then be equipped into the artificial retina system for filtering multiple primary objects to enhance vision. The suitability of handling multi-objects enables the system to cope with real-world complex scenarios. The proposed classification model is based on a multi-label deep neural network, specifically designed to leverage from the selective feature set. Initially, the enhanced images proposed in this research are compared with the ones that utilize an edge detection technique for single, dual, and multi-object images. These enhancements are also verified through an intensity profile analysis. Subsequently, the proposed classification model's performance is evaluated to show the significance of utilizing the suggested features. This includes evaluating the model's ability to correctly classify the top five, four, three, two, and one object(s), with respective accuracies of up to 84.8%, 85.2%, 86.8%, 91.8%, and 96.4%. Several comparisons such as training/validation loss and accuracies, precision, recall, specificity, and area under a curve indicate reliable results. Based on the overall evaluation of this study, it is concluded that using the suggested set of selective features not only improves the classification model's performance, but aligns with the specific problem to address the challenge of correctly identifying objects in multi-object scenarios. Therefore, the proposed classification model designed on the basis of selective features is considered to be a very useful tool in supporting the idea of optimizing image enhancement.

Asunto(s)

RESUMEN

Dynamic consent management allows a data subject to dynamically govern her consent to access her data. Clearly, security and privacy guarantees are vital for the adoption of dynamic consent management systems. In particular, specific data protection guarantees can be required to comply with rules and laws (e.g., the General Data Protection Regulation (GDPR)). Since the primary instantiation of the dynamic consent management systems in the existing literature is towards developing sustainable e-healthcare services, in this paper, we study data protection issues in dynamic consent management systems, identifying crucial security and privacy properties and discussing severe limitations of systems described in the state of the art. We have presented the precise definitions of security and privacy properties that are essential to confirm the robustness of the dynamic consent management systems against diverse adversaries. Finally, under those precise formal definitions of security and privacy, we have proposed the implications of state-of-the-art tools and technologies such as differential privacy, blockchain technologies, zero-knowledge proofs, and cryptographic procedures that can be used to build dynamic consent management systems that are secure and private by design.

RESUMEN

Data provenance means recording data origins and the history of data generation and processing. In healthcare, data provenance is one of the essential processes that make it possible to track the sources and reasons behind any problem with a user's data. With the emergence of the General Data Protection Regulation (GDPR), data provenance in healthcare systems should be implemented to give users more control over data. This SLR studies the impacts of data provenance in healthcare and GDPR-compliance-based data provenance through a systematic review of peer-reviewed articles. The SLR discusses the technologies used to achieve data provenance and various methodologies to achieve data provenance. We then explore different technologies that are applied in the healthcare domain and how they achieve data provenance. In the end, we have identified key research gaps followed by future research directions.

Asunto(s)

RESUMEN

This paper presents a 1600-pixel integrated neural stimulator with a correlated double-sampling readout (DSR) circuit for a subretinal prosthesis. The retinal stimulation chip inserted beneath the photoreceptor layer comprises an array of an active pixel sensor (APS) and biphasic pulse shaper. The DSR circuit achieves a high signal-to-noise ratio (SNR) of the APS with a short integration time to simultaneously improve the temporal and spatial resolutions of restored vision. This DSR circuit is adopted along with a 5 × 5-pixel tile, which reduces pixel size and improves the SNR by increasing the area occupied by storage capacitors. Moreover, a low-mismatch reference generator enables a low standard deviation between individual pulse shapers. The 1600-pixel retinal chip, fabricated using the 0.18 µm 1P6M CMOS process, occupies a total area of 4.3 mm × 3.3 mm and dissipates an average power of 3.4 mW; this was demonstrated by determining the stimulus current patterns corresponding to the illuminations of an LCD projector. Experimental results show that the proposed high-density stimulation array chip can achieve a high temporal resolution owing to its short integration time.

Asunto(s)

RESUMEN

Viruses infect millions of people worldwide each year, and some can lead to cancer or increase the risk of cancer. As viruses have highly mutable genomes, new viruses may emerge in the future, such as COVID-19 and influenza. Traditional virology relies on predefined rules to identify viruses, but new viruses may be completely or partially divergent from the reference genome, rendering statistical methods and similarity calculations insufficient for all genome sequences. Identifying DNA/RNA-based viral sequences is a crucial step in differentiating different types of lethal pathogens, including their variants and strains. While various tools in bioinformatics can align them, expert biologists are required to interpret the results. Computational virology is a scientific field that studies viruses, their origins, and drug discovery, where machine learning plays a crucial role in extracting domain- and task-specific features to tackle this challenge. This paper proposes a genome analysis system that uses advanced deep learning to identify dozens of viruses. The system uses nucleotide sequences from the NCBI GenBank database and a BERT tokenizer to extract features from the sequences by breaking them down into tokens. We also generated synthetic data for viruses with small sample sizes. The proposed system has two components: a scratch BERT architecture specifically designed for DNA analysis, which is used to learn the next codons unsupervised, and a classifier that identifies important features and understands the relationship between genotype and phenotype. Our system achieved an accuracy of 97.69% in identifying viral sequences.

RESUMEN

The present study aimed to evaluate the performance of automated skeletal maturation assessment system for Fishman's skeletal maturity indicators (SMI) for the use in dental fields. Skeletal maturity is particularly important in orthodontics for the determination of treatment timing and method. SMI is widely used for this purpose, as it is less time-consuming and practical in clinical use compared to other methods. Thus, the existing automated skeletal age assessment system based on Greulich and Pyle and Tanner-Whitehouse3 methods was further developed to include SMI using artificial intelligence. This hybrid SMI-modified system consists of three major steps: (1) automated detection of region of interest; (2) automated evaluation of skeletal maturity of each region; and (3) SMI stage mapping. The primary validation was carried out using a dataset of 2593 hand-wrist radiographs, and the SMI mapping algorithm was adjusted accordingly. The performance of the final system was evaluated on a test dataset of 711 hand-wrist radiographs from a different institution. The system achieved a prediction accuracy of 0.772 and mean absolute error and root mean square error of 0.27 and 0.604, respectively, indicating a clinically reliable performance. Thus, it can be used to improve clinical efficiency and reproducibility of SMI prediction.

Asunto(s)

RESUMEN

Introduction: The strut iliac bone graft has been widely used to achieve fusion in various anterior cervical spinal surgeries but some complications often remain, such as pain and gross deformity. Considering these, we designed a new technique to restore the iliac ridge, using the outmost part of the iliac crest. We aim to assess the efficacy of our new restoration technique of the iliac ridge after harvesting strut bone graft for anterior cervical fusion. The clinical and radiological outcomes of our hinged roof method were evaluated. Technical Note: A retrospective review was conducted of 29 patients who underwent hinged roof reconstruction of the iliac ridge after harvesting a bicortical strut bone graft for anterior cervical fusion using a cervical plate system. The clinical outcome for pain and gross appearance and radiological results were evaluated. Three months after the surgery, pain at the donor site became minimal or absent in all cases. At 1 year follow-up, no patient had reported pain and palpable discomfort, such as step-off on the donor site. Final X-ray and follow-up computed tomography revealed a bony union of the reconstructed iliac ridge to both margins. Conclusions: By showing good clinical and radiological outcomes, the authors' hinged roof reconstruction of the iliac crest after harvesting strut bone graft seemed to be a simple and effective technique that can reduce complications, such as pain and deformity on the donor iliac crest.

RESUMEN

This study examined the psychometric properties of the Barriers Self-Efficacy Scale-Physical Activity for Korean-speaking adults with osteoarthritis at risk for metabolic syndrome (N = 150). Factor analysis identified three dimensions of the Korean Barriers scale, explaining 65.9 % of the total variance. Confirmatory factor analysis indicated that the structural validity adequately fits the data. Construct validity confirmed significant associations between the amount of physical activity and psychological variables. The test-retest reliability was 0.87; the alpha was 0.90. The standardized response mean (0.497) indicated responsiveness to medium-magnitude change. The Korean Barriers scale can assess self-efficacy to engage in regular physical activity in clinical settings.

Asunto(s)

RESUMEN

Sensor technologies (including electrodes) have been widely utilized in many applications, especially in fields such as smart factories, automation, clinics, laboratories, and more [...].

Asunto(s)

RESUMEN

In this study, a pulse frequency modulation (PFM)-based stimulator is proposed for use in biomedical implantable devices. Conventionally, functional electrical stimulation (FES) techniques have been used to reinforce damaged nerves, such as retina tissue and brain tissue, by injecting a certain amount of charge into tissues. Although several design methods are present for implementing FES devices, an FES stimulator for retinal implants is difficult to realize because of the chip area, which needs to be inserted in a fovea, sized 5 mm x 5 mm, and power limitations to prevent the heat generation that causes tissue damage. In this work, we propose a novel stimulation structure to reduce the compliance voltage during stimulation, which can result in high-speed and low-voltage operation. A new stimulator that is composed of a modified high-speed PFM, a 4-bit counter, a serializer, a digital controller, and a current driver is designed and verified using a DB HiTek standard 0.18 µm process. This proposed stimulator can generate a charge up to 130 nC, consumes an average power of 375 µW during a stimulation period, and occupies a total area of 700 µm × 68 µm.

Asunto(s)

RESUMEN

COVID-19 has had a disproportionate impact on the elderly, who are over-represented among those who suffered severe illness or death. The obvious implication is that the share of the elderly in the population significantly affects the impact of COVID-19 on the overall health of a country. More generally, the elderly share has far-reaching economic and social ramifications. In this paper, we perform empirical analysis of cross-country data from 1970 to 2018 to identify the determinants of the share of the elderly-i.e., those aged 65 and over-in a country's population. We find that the quality of health care, life expectancy, and female labor participation increases the elderly share while higher fertility and female education attainment lower the elderly share. In addition, we find that the share is higher for high income countries and countries in Europe and Central Asia.

RESUMEN

Voice-activated artificial intelligence (AI) technology has advanced rapidly and is being adopted in various devices such as smart speakers and display products, which enable users to multitask without touching the devices. However, most devices equipped with cameras and displays lack mobility; therefore, users cannot avoid touching them for face-to-face interactions, which contradicts the voice-activated AI philosophy. In this paper, we propose a deep neural network-based real-time sound source localization (SSL) model for low-power internet of things (IoT) devices based on microphone arrays and present a prototype implemented on actual IoT devices. The proposed SSL model delivers multi-channel acoustic data to parallel convolutional neural network layers in the form of multiple streams to capture the unique delay patterns for the low-, mid-, and high-frequency ranges, and estimates the fine and coarse location of voices. The model adapted in this study achieved an accuracy of 91.41% on fine location estimation and a direction of arrival error of 7.43° on noisy data. It achieved a processing time of 7.811 ms per 40 ms samples on the Raspberry Pi 4B. The proposed model can be applied to a camera-based humanoid robot that mimics the manner in which humans react to trigger voices in crowded environments.

Asunto(s)

RESUMEN

Significant progress has been made in the field of micro/nano-retinal implant technologies. However, the high pixel range, power leakage, reliability, and lifespan of retinal implants are still questionable. Active implantable devices are safe, cost-effective, and reliable. Although a device that can meet basic safety requirements set by the Food and Drug Administration and the European Union is reliable for long-term use and provides control on current and voltage parameters, it will be expensive and cannot be commercially successful. This study proposes an economical, fully controllable, and configurable wireless communication system based on field-programmable gated arrays (FPGAs) that were designed with the ability to cope with the issues that arise in retinal implantation. This system incorporates hexagonal biphasic stimulation pulses generated by a digital controller that can be fully controlled using an external transmitter. The integration of two separate domain analog systems and a digital controller based on FPGAs is proposed in this study. The system was also implemented on a microchip and verified using in vitro results.

Asunto(s)

RESUMEN

Power-efficient digital controllers are proposed for wireless retinal prosthetic systems. Power management plays an important role in reducing the power consumption and avoiding malfunctions in implantable medical devices. In the case of implantable devices with only one-way communication, the received power level is uncertain because there is no feedback on the power status. Accordingly, system breakdown due to inefficient power management should be avoided to prevent harm to patients. In this study, digital power controllers were developed for achieving two-way communication. Three controllers-a forward and back telemetry control unit, a power control unit, and a preamble control unit-operated simultaneously to control the class-E amplifier input power, provided command data to stimulators, monitored the power levels of the implanted devices, and generated back telemetry data. For performance verification, we implemented a digital power control system using a field-programmable gate array and then demonstrated it by employing a wireless telemetry system.

Asunto(s)

RESUMEN

In this study, we propose a low-area multi-channel controlled dielectric breakdown (CDB) system that simultaneously produces several nanopore sensors. Conventionally, solid-state nanopores are prepared by etching or drilling openings in a silicon nitride (SiNx) substrate, which is expensive and requires a long processing time. To address these challenges, a CDB technique was introduced and used to fabricate nanopore channels in SiNx membranes. However, the nanopore sensors produced by the CDB result in a severe pore-to-pore diameter variation as a result of different fabrication conditions and processing times. Accordingly, it is indispensable to simultaneously fabricate nanopore sensors in the same environment to reduce the deleterious effects of pore-to-pore variation. In this study, we propose a four-channel CDB system that comprises an amplifier that boosts the command voltage, a 1-to-4 multiplexer, a level shifter, a low-noise transimpedance amplifier and a data acquisition device. To prove our design concept, we used the CDB system to fabricate four nanopore sensors with diameters of <10 nm, and its in vitro performance was verified using λ-DNA samples.

Asunto(s)

RESUMEN

This study examined the psychometric properties of the Korean version of the Patient Knowledge Questionnaire-Osteoarthritis (PKQ-OA-K). A cross-sectional survey was conducted with 157 adults with osteoarthritis from the outpatient clinic at a university hospital in Korea. The overall correct answer rate for the PKQ-OA-K was 60.4%; notably, the drug therapy subscale had the lowest median score percentage (42.9%). For structural validity, exploratory factor analysis identified the PKQ-OA-K as two-dimensional, explaining 52.4% of the total variance. Confirmatory factor analysis showed that the two-factor model adequately fit the data. The PKQ-OA-K was positively correlated with education level (r = 0.24) and osteoarthritis outcomes (r = 0.17), thus verifying the hypotheses of construct validity. The intraclass correlation coefficient for test-retest reliability was 0.52; alpha was 0.44. The PKQ-OA-K has excellent validity but imperfect reliability for adults with osteoarthritis. This study recommends cautious use of the PKQ-OA-K to assess Korean patients' knowledge of osteoarthritis.

Asunto(s)

RESUMEN

The COVID-19 pandemic has given rise to a spike in financial market volatility. In this paper, we attempt to assess the effects of financial & news-driven uncertainty shocks in growing Asian economies, using country-specific bond volatility shocks as a measure of local interest rate uncertainty. Also, we contrast the effects of local uncertainty with global stock market uncertainty. Using bond market data from nine Asian markets, we uncover a transmission mechanism of uncertainty shocks via the bond market. The mechanism works as a crowding-out effect due to government-led excessive market borrowing with supply-side consequences for the private sector, as opposed to economic policy or global stock market uncertainty which works more like a demand shock as in the literature. We conclude that countries with growing fiscal deficits that entail a larger government bond market or higher current account deficits, tend to experience an increase in the cost of borrowing due to this bond market volatility or interest rate uncertainty shocks.