TinyFallNet: A Lightweight Pre-Impact Fall Detection Model.

Falls represent a significant health concern for the elderly. While studies on deep learning-based preimpact fall detection have been conducted to mitigate fall-related injuries, additional efforts are needed for embedding in microcomputer units (MCUs). In this study, ConvLSTM, the state-of-the-art model, was benchmarked, and we attempted to lightweight it by leveraging features from image-classification models VGGNet and ResNet while maintaining performance for wearable airbags. The models were developed and evaluated using data from young subjects in the KFall public dataset based on an inertial measurement unit (IMU), leading to the proposal of TinyFallNet based on ResNet. Despite exhibiting higher accuracy (97.37% < 98.00%) than the benchmarked ConvLSTM, the proposed model requires lower memory (1.58 MB > 0.70 MB). Additionally, data on the elderly from the fall data of the FARSEEING dataset and activities of daily living (ADLs) data of the KFall dataset were analyzed for algorithm validation. This study demonstrated the applicability of image-classification models to preimpact fall detection using IMU and showed that additional tuning for lightweighting is possible due to the different data types. This research is expected to contribute to the lightweighting of deep learning models based on IMU and the development of applications based on IMU data.

Development of a toroidally resolved broadband ECE imaging system for measurement of turbulent fluctuations on the KSTAR.

The two electron cyclotron emission imaging (ECEI) systems installed at adjacent ports (G and H) on the KSTAR tokamak incorporate large-aperture mm-wave optics, broadband electronics, and high speed digitization (up to 1 MSa/s) for 2D and quasi-3D visualization of MHD-scale fluid dynamics. Recently, the ECEI systems have been proved to be capable of visualization of smaller scale fluctuations albeit with a limited spatiotemporal resolution and even capable of measurement of ion cyclotron harmonic waves by direct high-speed sampling of the ECE IF signals. A four-channel prototype subsystem with a higher sampling rate up to 16 GS/s has been integrated into the G-port ECEI system, enabling the measurement of plasma waves in the GHz range in the form of modulated ECE signals and characterization of high-frequency turbulence during the evolution of pedestal. To achieve higher toroidal resolution in the turbulence measurement, the H-port ECEI system is now being upgraded to have a toroidally dual detector array of 2(toroidal) × 12(vertical) × 8(radial) channel configuration and a high-speed subsystem of 2(toroidal) × 4 channel configuration. The new mm-wave optics has been designed via beam propagation simulation, and the measured performance of the fabricated lens indicates a toroidal resolution of 8-10 cm depending on the focus position and zoom factor, allowing for the measurement of parallel wavenumber up to k‖ ∼ 0.8 cm-1.

Numerical Study on Effective Conditions for the Induction of Apoptotic Temperatures for Various Tumor Aspect Ratios Using a Single Continuous-Wave Laser in Photothermal Therapy Using Gold Nanorods.

Photothermal therapy can serve as an alternative to classic surgery in the treatment of patients with cancer. However, using photothermal therapy can result in local overheating and damage to normal tissues. Therefore, it is important to determine effective heating conditions based on heat transfer. In this study, we analyzed laser-tissue interactions in gold nanoparticle (GNP)-enhanced photothermal therapy based on the theory of heat transfer. The thermal behavior inside tissues during photothermal therapy was analyzed using numerical analysis. The apoptosis ratio was defined by deriving the area having a temperature distribution between 43 °C and 50 °C, which is required for inducing apoptosis. Thermal damage, caused by local heating, was defined using the thermal hazard value. Using this approach, we confirmed that apoptosis can be predicted with respect to tumor size (aspect ratio) and heating conditions (laser intensity and radius) in photothermal therapy with a continuous-wave laser. Finally, we determined the effective apoptosis ratio and thermal hazard value of normal tissue according to tumor size and heating conditions, thereby establishing conditions for inducing maximal levels of cell apoptosis with minimal damage to normal tissue. The optimization conditions proposed in this study can be a gentle and effective treatment option for photothermal therapy.

The Morphological Effects of ZnO Upon the Antimicrobial and Deodorant Activities of Polyethylene Terephthalate/ZnO Composite Filaments.

In this study, the morphological effects of ZnO on the antimicrobial and deodorant activities of synthetic fibers were investigated. Three different polyethylene terephthalate (PET)/ZnO filaments were prepared by incorporating various ZnO nanostructures (rods, plates, and spheres) into PET filaments via a melt-spinning process. The antimicrobial activity of the as-prepared fibers was evaluated by the shake-flask method using two types of bacteria (Staphylococcus aureus and Klebsiella pneumoniae). The deodorant activity of the as-prepared fibers was evaluated by the gas detection tube method. All the PET/ZnO filaments exhibited excellent antimicrobial activity with a bacterial reduction value of 99.9%. The PET/ZnO rod filament showed the best deodorant performance of 60.0%. Both the antimicrobial and deodorant activities of the PET/ZnO filaments were influenced by the morphology of ZnO. However, the morphology of ZnO had a different effect on each functionality of the PET/ZnO filaments. The antimicrobial activity of the PET/ZnO filaments was mainly affected by the physical properties of ZnO rather than its morphology. By contrast, the deodorant activity of the PET/ZnO filaments was highly influenced by the morphology of ZnO.

Label-Free Electrical Detection of Enzymatic Reactions in Nanochannels.

We report label-free electrical detection of enzymatic reactions using 2-D nanofluidic channels and investigate reaction kinetics of enzymatic reactions on immobilized substrates in nanoscale-confined spaces. Trypsin proteolysis is chosen for demonstration of the detection scheme. When trypsin cleaves poly-l-lysine coated on the surface of silica nanochannels, the resulting change of surface charge density can be detected by monitoring the ionic conductance of the nanochannels. Our results show that detection of such surface enzymatic reactions is faster than detection of surface binding reactions in nanochannels for low-concentration analytes. Furthermore, the nanochannel sensor has a sensitivity down to 5 ng/mL, which statistically corresponds to a single enzyme per nanochannel. Our results also suggest that enzyme kinetics in nanochannels is fundamentally different from that in bulk solutions or plain surfaces. Such enzymatic reactions form two clear self-propagating reaction fronts inside the nanochannels, and the reaction fronts follow square-root time dependences at high enzyme concentrations due to significant nonspecific adsorption. However, at low enzyme concentrations when nonspecific adsorption is negligible, the reaction fronts propagate linearly with time, and the corresponding propagation speed is related to the channel geometry, enzyme concentration, catalytic reaction constant, diffusion coefficient, and substrate surface density. Optimization of this nanochannel sensor could lead to a quick-response, highly sensitive, and label-free sensor for enzyme assay and kinetic studies.

Comparison of surface properties of random, block, and graft copolymers having perfluoroalkyl and silicone-containing side chains.

Fluorosilicone copolymers of random, block, and graft with both perfluoroalkyl and silicone-containing side chains were synthesized, and their surface properties and surface modification effects on PVC film were compared. It can be confirmed that the fluorosilicone copolymers of random, block, and graft exhibit very low surface free energies of 9-13 dyn/cm, depending on the perfluoroalkyl group content and their molecular structure. The inherent surface free energies of the fluorosilicone copolymers are significantly influenced by their molecular structure and perfluoroalkyl group content. It can also be found that the fluorosilicone copolymers are very effective for lowering surface free energy. The surface free energy of a copolymer/PVC blend strongly varies with perfluoroalkyl group content as well as molecular structure. The molecular structure of a fluorosilicone copolymer is as important as the perfluoroalkyl group content for their inherent surface free energies and surface modification of other polymers.

Scanning flow-impedance microscopy: a simple imaging technique based on hydrodynamics.

We introduce the concept of scanning flow-impedance microscopy (SFIM) which is an imaging technique based on hydrodynamics. Using a simple experimental setup including a mass flow controller and a manometer, the operating principle of SFIM is validated under atmospheric pressure and temperature conditions. Experimental results show that the flow impedance strongly depends on the relative distance between a probe and a specimen. SFIM micrographs of microscale patterns with various linewidths are presented.

Extremely uniform angular distributions of the three-dimensional emission spectra of photonic quantum ring lasers.

We have observed the emission spectra of three-dimensional Rayleigh-Fabry-Perot cavity whispering-gallery modes from photonic quantum ring lasers that give rise to uniform angular mode spacing in space. The observed angular distribution (0 degrees approximately 30 degrees) of the emission modes is linear with respect to the mode number. For example, the average angular mode spacing is 0.95 degrees with a standard deviation of 0.098 degrees for a 20-microm-diameter laser, whose potential angle-resolving power is promising for angle-sensing applications in the future.

Preparation of Water-Repellent Glass by Sol-Gel Process Using Perfluoroalkylsilane and Tetraethoxysilane.

Coating films on glass substrate were prepared by sol-gel process using alkoxide solutions containing perfluoroalkylsilane (PFAS) and tetraethoxysilane (TEOS). The physical properties of the coating films were characterized by SEM, FT-IR, and XRD. And their surface properties were investigated by measuring contact angles and atomic compositions. Transparent coating films with smooth surface and uniform thickness could be obtained. The contact angles of the coating films for water and methylene iodide are extremely high, at 118 degrees and 97 degrees, respectively, and their surface free energies are about 9.7 dyn/cm. It was found that the water-repellent glass prepared is very hydrophobic and exhibits excellent water-repellency. Hydrophobic perfluoroalkyl groups are preferentially enriched to the outermost layer at the coating film-air interface, and two layers probably exist in the coating film. The upper layer oriented toward the air is composed of mainly perfluoroalkyl groups originating from PFAS, and the lower layer is composed of mainly -OSiO- groups originating from TEOS. The heat treatment after drying step cannot influence the surface enrichment of the perfluoroalkyl group. The hydrolysis reaction should be more completely done before the dip coating step to obtain lower surface free energy. The burning temperature should be less than 300 degrees C because the perfluoroalkyl group begins to decompose from this temperature. Copyright 2001 Academic Press.