Novel approach exploring the correlation between presepsin and routine laboratory parameters using explainable artificial intelligence.

Although presepsin, a crucial biomarker for the diagnosis and management of sepsis, has gained prominence in contemporary medical research, its relationship with routine laboratory parameters, including demographic data and hospital blood test data, remains underexplored. This study integrates machine learning with explainable artificial intelligence (XAI) to provide insights into the relationship between presepsin and these parameters. Advanced machine learning classifiers provide a multilateral view of data and play an important role in highlighting the interrelationships between presepsin and other parameters. XAI enhances analysis by ensuring transparency in the model's decisions, especially in selecting key parameters that significantly enhance classification accuracy. Utilizing XAI, this study successfully identified critical parameters that increased the predictive accuracy for sepsis patients, achieving a remarkable ROC AUC of 0.97 and an accuracy of 0.94. This breakthrough is possibly attributed to the comprehensive utilization of XAI in refining parameter selection, thus leading to these significant predictive metrics. The presence of missing data in datasets is another concern; this study addresses it by employing Extreme Gradient Boosting (XGBoost) to manage missing data, effectively mitigating potential biases while preserving both the accuracy and relevance of the results. The perspective of examining data from higher dimensions using machine learning transcends traditional observation and analysis. The findings of this study hold the potential to enhance patient diagnoses and treatment, underscoring the value of merging traditional research methods with advanced analytical tools.

Column Row Convolutional Neural Network: Reducing Parameters for Efficient Image Processing.

Recent advancements in deep learning have achieved significant progress by increasing the number of parameters in a given model. However, this comes at the cost of computing resources, prompting researchers to explore model compression techniques that reduce the number of parameters while maintaining or even improving performance. Convolutional neural networks (CNN) have been recognized as more efficient and effective than fully connected (FC) networks. We propose a column row convolutional neural network (CRCNN) in this letter that applies 1D convolution to image data, significantly reducing the number of learning parameters and operational steps. The CRCNN uses column and row local receptive fields to perform data abstraction, concatenating each direction's feature before connecting it to an FC layer. Experimental results demonstrate that the CRCNN maintains comparable accuracy while reducing the number of parameters and compared to prior work. Moreover, the CRCNN is employed for one-class anomaly detection, demonstrating its feasibility for various applications.

Purely self-rectifying memristor-based passive crossbar array for artificial neural network accelerators.

Memristor-integrated passive crossbar arrays (CAs) could potentially accelerate neural network (NN) computations, but studies on these devices are limited to software-based simulations owing to their poor reliability. Herein, we propose a self-rectifying memristor-based 1 kb CA as a hardware accelerator for NN computations. We conducted fully hardware-based single-layer NN classification tasks involving the Modified National Institute of Standards and Technology database using the developed passive CA, and achieved 100% classification accuracy for 1500 test sets. We also investigated the influences of the defect-tolerance capability of the CA, impact of the conductance range of the integrated memristors, and presence or absence of selection functionality in the integrated memristors on the image classification tasks. We offer valuable insights into the behavior and performance of CA devices under various conditions and provide evidence of the practicality of memristor-integrated passive CAs as hardware accelerators for NN applications.

Fibriform Organic Electrochemical Diodes with Rectifying, Complementary Logic and Transient Voltage Suppression Functions for Wearable E-Textile Embedded Circuits.

In this study, a fibriform electrochemical diode capable of performing rectifying, complementary logic and device protection functions for future e-textile circuit systems is fabricated. The diode was fabricated using a simple twisted assembly of metal/polymer semiconductor/ion gel coaxial microfibers and conducting microfiber electrodes. The fibriform diode exhibited a prominent asymmetrical current flow with a rectification ratio of over 102, and its performance was retained after repeated bending deformations and washings. Fundamental studies on the electrochemical interactions of polymer semiconductors with ions reveal that the Faradaic current generated in polymer semiconductors by electrochemical reactions results in an abrupt current increase under a forward bias, in which the threshold voltages of the device are determined by the oxidation or reduction potential of the polymer semiconductor. Textile-embedded full-wave rectifiers and logic gate circuits were implemented by simply integrating the fibriform diodes, exhibiting AC-to-DC signal conversion and logic operation functions, respectively. It was also confirmed that the proposed fibriform diode can suppress transient voltages and thus protect a low-voltage operational wearable e-textile circuit.

Bio-inspired electronic fingerprint PUF device with single-walled carbon nanotube network surface mediated by M13 bacteriophage template.

Human fingerprints are randomly created during fetal activity in the womb, resulting in unique and physically irreproducible fingerprint patterns that are applicable as a biological cryptographic primitive. Similarly, stochastically knitted single-walled carbon nanotube (SWNT) network surfaces exhibit inherently random and unique electrical characteristics that can be exploited as a physical unclonable function (PUF) in the authentication. In this study, filamentous M13 bacteriophages are used as a biological gluing template to create a random SWNT network surface with mechanical flexibility, with electrical properties determined by random variation during fabrication. The resistance profile between two adjacent electrodes was mapped for these M13-mediated SWNT network surfaces, with the results demonstrating a unique resistance profile for each M13-SWNT device, similar to that of human fingerprints. Randomness and uniqueness measures were evaluated as respectively 50.5% and 50% using generated challenge-response pairs. Min-entropy for unpredictability evaluation of the M13-SWNT based PUFs resulted in 0.98. Our results showed that M13-SWNT random network exhibits cryptographic characteristics when used in a bio-inspired PUF device.

Dendritic Network Implementable Organic Neurofiber Transistors with Enhanced Memory Cyclic Endurance for Spatiotemporal Iterative Learning.

Dendritic network implementable organic neurofiber transistors with enhanced memory cyclic endurance for spatiotemporal iterative learning are proposed. The architecture of the fibrous organic electrochemical transistors consisting of a double-stranded assembly of electrode microfibers and an iongel gate insulator enables the highly sensitive multiple implementation of synaptic junctions via simple physical contact of gate-electrode microfibers, similar to the dendritic connections of a biological neuron fiber. In particular, carboxylic-acid-functionalized polythiophene as a semiconductor channel material provides stable gate-field-dependent multilevel memory characteristics with long-term stability and cyclic endurance, unlike the conventional poly(alkylthiophene)-based neuromorphic electrochemical transistors, which exhibit short retention and unstable endurance. The dissociation of the carboxylic acid of the polythiophene enables reversible doping and dedoping of the polythiophene channel by effectively stabilizing the ions that penetrate the channel during potentiation and depression cycles, leading to the reliable cyclic endurance of the device. The synaptic weight of the neurofiber transistors with a dendritic network maintains the state levels stably and is independently updated with each synapse connected with the presynaptic neuron to a specific state level. Finally, the neurofiber transistor demonstrates successful speech recognition based on iterative spiking neural network learning in the time domain, showing a substantial recognition accuracy of 88.9%.

Distribution of petrogenic polycyclic aromatic hydrocarbons (PAHs) in seafood following Deepwater Horizon oil spill.

A community-based participatory research was utilized to address the coastal community's concern regarding Deepwater Horizon oil contamination of seafood. Therefore, we analyzed polycyclic aromatic hydrocarbons (PAHs), major toxic constituents of crude oil, in the seafood collected from gulf coast (Louisiana, Alabama and Mississippi) during December 2011-February 2014. PAHs were extracted from edible part of shrimp, oysters, and crabs by the QuEChERS/dsPE procedure and analyzed by gas chromatography-mass spectrometry. The total PAHs data were further analyzed using the General Linear Mixed Model procedure of the SAS (Version 9.3, SAS Institute, Inc., Cary, NC) statistical software. Brown shrimp showed statistically significant differences in PAHs levels with respect to time and locations while white shrimp showed differences at various time points. PAHs levels in oyster and crab samples were not statistically different at the Type I error of 0.05. Overall, the PAHs levels are far below FDA levels of concern for human consumption.

A New Architecture for Fibrous Organic Transistors Based on a Double-Stranded Assembly of Electrode Microfibers for Electronic Textile Applications.

Herein, a unique device architecture is proposed for fibrous organic transistors based on a double-stranded assembly of electrode microfibers for electronic textile applications. A key feature of this work is that the semiconductor channel of the fiber transistor comprises a twist assembly of the source and drain electrode microfibers that are coated by an organic semiconductor. This architecture not only allows the channel dimension of the device to be readily controlled by varying the thickness of the semiconductor layer and the twisted length of the two electrode microfibers, but also passivates the device without affecting interconnections with other electrical components. It is found that the control of crystalline nanostructure of the semiconductor layer is critical for improving both the production yield of the device and the charge-carrier transport in the device. The resulting fibrous organic transistors show a high output current of over -5 mA at a low operation voltage of -1.3 V and a good on/off current ratio of 105 . The device performance is maintained after repeated bending deformation and washing with a strong detergent solution. Application of the fibrous organic transistors to switch current-driven LED devices and detection of electrocardiography signals from a human body are demonstrated.

How do plants see the world? - UV imaging with a TiO2 nanowire array by artificial photosynthesis.

The concept of plant vision refers to the fact that plants are receptive to their visual environment, although the mechanism involved is quite distinct from the human visual system. The mechanism in plants is not well understood and has yet to be fully investigated. In this work, we have exploited the properties of TiO2 nanowires as a UV sensor to simulate the phenomenon of photosynthesis in order to come one step closer to understanding how plants see the world. To the best of our knowledge, this study is the first approach to emulate and depict plant vision. We have emulated the visual map perceived by plants with a single-pixel imaging system combined with a mechanical scanner. The image acquisition has been demonstrated for several electrolyte environments, in both transmissive and reflective configurations, in order to explore the different conditions in which plants perceive light.

Four-Bits-Per-Cell Operation in an HfO2 -Based Resistive Switching Device.

The quadruple-level cell technology is demonstrated in an Au/Al2 O3 /HfO2 /TiN resistance switching memory device using the industry-standard incremental step pulse programming (ISPP) and error checking/correction (ECC) methods. With the highly optimistic properties of the tested device, such as self-compliance and gradual set-switching behaviors, the device shows 6σ reliability up to 16 states with a state current gap value of 400 nA for the total allowable programmed current range from 2 to 11 µA. It is demonstrated that the conventional ISPP/ECC can be applied to such resistance switching memory, which may greatly contribute to the commercialization of the device, especially competitively with NAND flash. A relatively minor improvement in the material and circuitry may enable even a five-bits-per-cell technology, which can hardly be imagined in NAND flash, whose state-of-the-art multiple-cell technology is only at three-level (eight states) to this day.