ECG Biometric Authentication Using Self-Supervised Learning for IoT Edge Sensors.

Wearable Internet of Things (IoT) devices are gaining ground for continuous physiological data acquisition and health monitoring. These physiological signals can be used for security applications to achieve continuous authentication and user convenience due to passive data acquisition. This paper investigates an electrocardiogram (ECG) based biometric user authentication system using features derived from the Convolutional Neural Network (CNN) and self-supervised contrastive learning. Contrastive learning enables us to use large unlabeled datasets to train the model and establish its generalizability. We propose approaches enabling the CNN encoder to extract appropriate features that distinguish the user from other subjects. When evaluated using the PTB ECG database with 290 subjects, the proposed technique achieved an authentication accuracy of 99.15%. To test its generalizability, we applied the model to two new datasets, the MIT-BIH Arrhythmia Database and the ECG-ID Database, achieving over 98.5% accuracy without any modifications. Furthermore, we show that repeating the authentication step three times can increase accuracy to nearly 100% for both PTBDB and ECGIDDB. This paper also presents model optimizations for embedded device deployment, which makes the system more relevant to real-world scenarios. To deploy our model in IoT edge sensors, we optimized the model complexity by applying quantization and pruning. The optimized model achieves 98.67% accuracy on PTBDB, with 0.48% accuracy loss and 62.6% CPU cycles compared to the unoptimized model. An accuracy-vs-time-complexity tradeoff analysis is performed, and results are presented for different optimization levels.

Synthesis of cholesterol analogues and comparison on their effect on plasma cholesterol with ß-Sitosterol.

The application of plant sterols in the treatment of hypercholesterolemia is promising. We hypothesize that plant sterols can reduce blood cholesterol because they have a side chain of at least three branches. Three cholesterol analogues were synthesized: CA0 (no side chain), CA3 (a 3­carbon chain with one branch), and CA14 (a 14­carbon side chain with two branches), and then compared their effect on blood cholesterol with that of ß-sitosterol. Structurally, ß-sitosterol has a 10­carbon side chain with three branches. Results demonstrated that ß-sitosterol could effectively reduce plasma total cholesterol (TC) by 20.3%, whereas CA0, CA3 and CA14 did not affect plasma TC in hypercholesterolemia hamsters. ß-Sitosterol was absent in the plasma and liver, indicating it was not absorbed. We concluded that ß-sitosterol with three branches had plasma TC-lowering activity. In contrast, cholesterol analogues with a side chain of two or fewer branches did not affect plasma cholesterol.

Dominating Alzheimer's disease diagnosis with deep learning on sMRI and DTI-MD.

Background: Alzheimer's disease (AD) is a progressive and irreversible neurodegenerative disorder that has become one of the major health concerns for the elderly. Computer-aided AD diagnosis can assist doctors in quickly and accurately determining patients' severity and affected regions. Methods: In this paper, we propose a method called MADNet for computer-aided AD diagnosis using multimodal datasets. The method selects ResNet-10 as the backbone network, with dual-branch parallel extraction of discriminative features for AD classification. It incorporates long-range dependencies modeling using attention scores in the decision-making layer and fuses the features based on their importance across modalities. To validate the effectiveness of our proposed multimodal classification method, we construct a multimodal dataset based on the publicly available ADNI dataset and a collected XWNI dataset, which includes examples of AD, Mild Cognitive Impairment (MCI), and Cognitively Normal (CN). Results: On this dataset, we conduct binary classification experiments of AD vs. CN and MCI vs. CN, and demonstrate that our proposed method outperforms other traditional single-modal deep learning models. Furthermore, this conclusion also confirms the necessity of using multimodal sMRI and DTI data for computer-aided AD diagnosis, as these two modalities complement and convey information to each other. We visualize the feature maps extracted by MADNet using Grad-CAM, generating heatmaps that guide doctors' attention to important regions in patients' sMRI, which play a crucial role in the development of AD, establishing trust between human experts and machine learning models. Conclusion: We propose a simple yet effective multimodal deep convolutional neural network model MADNet that outperforms traditional deep learning methods that use a single-modality dataset for AD diagnosis.

A phytic acid-based CoMn bimetallic metal-organic aerogel (PA-CoMn-1 : x : 2 - x) as a modifier for electrochemical detection of non-steroidal anti-inflammatory drug 4-acetaminophenol.

Metal phosphides and phosphate-based materials have received significant attention due to their high electrocatalytic activity, adjustable structure composition and stability. Herein, we introduce a phytic acid-based CoMn bimetallic metal-organic (PA-CoMn) aerogel as an electrode modifier, derived from PA and mixed transition metal ions (Co2+, Mn2+). We explored its performance in the sensitive sensing of non-steroidal anti-inflammatory drug 4-acetaminophenol (4-AP) for the first time. We investigated the electrochemical behavior of the modified screen-printed carbon electrode (SPCE) by cyclic voltammetry (CV) and differential pulse voltammetry (DPV). The PA-CoMn-1 : 1.5 : 0.5 aerogel/Nafion/SPCE proved to be highly sensitive and selective towards the detection of 4-AP. A double linear response was recorded for 4-AP over the range of 1 µM to 0.1 mM and lower detection limits (LOD) of 0.2133 µM. The applicability of the PA-CoMn-1 : 1.5 : 0.5 aerogel/Nafion/SPCE in the detection of 4-AP in commercial drug samples with good recoveries was investigated, confirming the great potential of PA-CoMn-1 : 1.5 : 0.5 aerogel/SPCE in clinical applications.

A multifunctional gingival retraction cord with antibacterial and hemostasis properties based on Chitosan/Propolis/Tranexamic acid for dental treatment.

A novel gingival retraction cord named P/TA@CSy was prepared using chitosan yarns (CSy) loaded with tranexamic acid (TA) and Propolis (P). P/TA@CSy has good toughness with a breaking strength of 41.3 Pa, benefiting from the twisting structure and Propolis coating. A short coagulation time of 456 s was achieved for P/TA@CSy because of the potent blood absorption ability from the effective attachment of tranexamic acid. Moreover, excellent antibacterial ability was obtained with the antibacterial rates against E. coli of 94.73 %, S. aureus of 99.99 % and S. mutans of 99.99 %, contributing to Propolis's antibacterial ability. In addition, suppression of the expression of pro-inflammatory cytokines (IL-6 and TNF-α) was found, which could prevent wound infection. P/TA@CSy displayed excellent cytocompatibility with the cell activity of 100 % after 24 h. Therefore, P/TA@CSy could rapidly respond to gingival hemostasis and infection prevention, showing excellent potential in dental treatment.

Unlocking the NIR-II AIEgen for High Brightness through Intramolecular Electrostatic Locking.

Fluorescent imaging and biosensing in the near-infrared-II (NIR-II) window holds great promise for non-invasive, radiation-free, and rapid-response clinical diagnosis. However, it's still challenging to develop bright NIR-II fluorophores. In this study, we report a new strategy to enhance the brightness of NIR-II aggregation-induced emission (AIE) fluorophores through intramolecular electrostatic locking. By introducing sulfur atoms into the side chains of the thiophene bridge in TSEH molecule, the molecular motion of the conjugated backbone can be locked through intramolecular interactions between the sulfur and nitrogen atoms. This leads to enhanced NIR-II fluorescent emission of TSEH in both solution and aggregation states. Notably, the encapsulated nanoparticles (NPs) of TSEH show enhanced brightness, which is 2.6-fold higher than TEH NPs with alkyl side chains. The in vivo experiments reveal the feasibility of TSEH NPs in vascular and tumor imaging with a high signal-to-background ratio and precise resection for tiny tumors. In addition, polystyrene nanospheres encapsulated with TSEH are utilized for antigen detection in lateral flow assays, showing a signal-to-noise ratio 1.9-fold higher than the TEH counterpart in detecting low-concentration antigens. This work highlights the potential for developing bright NIR-II fluorophores through intramolecular electrostatic locking and their potential applications in clinical diagnosis and biomedical research.

Robust deep learning from incomplete annotation for accurate lung nodule detection.

Deep learning plays a significant role in the detection of pulmonary nodules in low-dose computed tomography (LDCT) scans, contributing to the diagnosis and treatment of lung cancer. Nevertheless, its effectiveness often relies on the availability of extensive, meticulously annotated dataset. In this paper, we explore the utilization of an incompletely annotated dataset for pulmonary nodules detection and introduce the FULFIL (Forecasting Uncompleted Labels For Inexpensive Lung nodule detection) algorithm as an innovative approach. By instructing annotators to label only the nodules they are most confident about, without requiring complete coverage, we can substantially reduce annotation costs. Nevertheless, this approach results in an incompletely annotated dataset, which presents challenges when training deep learning models. Within the FULFIL algorithm, we employ Graph Convolution Network (GCN) to discover the relationships between annotated and unannotated nodules for self-adaptively completing the annotation. Meanwhile, a teacher-student framework is employed for self-adaptive learning using the completed annotation dataset. Furthermore, we have designed a Dual-Views loss to leverage different data perspectives, aiding the model in acquiring robust features and enhancing generalization. We carried out experiments using the LUng Nodule Analysis (LUNA) dataset, achieving a sensitivity of 0.574 at a False positives per scan (FPs/scan) of 0.125 with only 10% instance-level annotations for nodules. This performance outperformed comparative methods by 7.00%. Experimental comparisons were conducted to evaluate the performance of our model and human experts on test dataset. The results demonstrate that our model can achieve a comparable level of performance to that of human experts. The comprehensive experimental results demonstrate that FULFIL can effectively leverage an incomplete pulmonary nodule dataset to develop a robust deep learning model, making it a promising tool for assisting in lung nodule detection.

Self-supervised contrastive learning for EEG-based cross-subject motor imagery recognition.

Objective. The extensive application of electroencephalography (EEG) in brain-computer interfaces (BCIs) can be attributed to its non-invasive nature and capability to offer high-resolution data. The acquisition of EEG signals is a straightforward process, but the datasets associated with these signals frequently exhibit data scarcity and require substantial resources for proper labeling. Furthermore, there is a significant limitation in the generalization performance of EEG models due to the substantial inter-individual variability observed in EEG signals.Approach. To address these issues, we propose a novel self-supervised contrastive learning framework for decoding motor imagery (MI) signals in cross-subject scenarios. Specifically, we design an encoder combining convolutional neural network and attention mechanism. In the contrastive learning training stage, the network undergoes training with the pretext task of data augmentation to minimize the distance between pairs of homologous transformations while simultaneously maximizing the distance between pairs of heterologous transformations. It enhances the amount of data utilized for training and improves the network's ability to extract deep features from original signals without relying on the true labels of the data.Main results. To evaluate our framework's efficacy, we conduct extensive experiments on three public MI datasets: BCI IV IIa, BCI IV IIb, and HGD datasets. The proposed method achieves cross-subject classification accuracies of 67.32%, 82.34%, and 81.13%on the three datasets, demonstrating superior performance compared to existing methods.Significance. Therefore, this method has great promise for improving the performance of cross-subject transfer learning in MI-based BCI systems.

Multi modality fusion transformer with spatio-temporal feature aggregation module for psychiatric disorder diagnosis.

Bipolar disorder (BD) is characterized by recurrent episodes of depression and mild mania. In this paper, to address the common issue of insufficient accuracy in existing methods and meet the requirements of clinical diagnosis, we propose a framework called Spatio-temporal Feature Fusion Transformer (STF2Former). It improves on our previous work - MFFormer by introducing a Spatio-temporal Feature Aggregation Module (STFAM) to learn the temporal and spatial features of rs-fMRI data. It promotes intra-modality attention and information fusion across different modalities. Specifically, this method decouples the temporal and spatial dimensions and designs two feature extraction modules for extracting temporal and spatial information separately. Extensive experiments demonstrate the effectiveness of our proposed STFAM in extracting features from rs-fMRI, and prove that our STF2Former can significantly outperform MFFormer and achieve much better results among other state-of-the-art methods.

Role of Chinese Acupuncture in the Treatment for Chemotherapy-Induced Cognitive Impairment in Older Patients With Cancer: Protocol for a Randomized Controlled Trial.

BACKGROUND: Older patients with cancer experience cognitive impairment and a series of neurocognitive symptoms known as chemobrain due to chemotherapy. Moreover, older populations are disproportionately affected by chemobrain and heightened negative mental health outcomes after cytotoxic chemical drug therapy. Chinese acupuncture is an emerging therapeutic option for chemotherapy-induced cognitive impairment in older patients with cancer, despite limited supporting evidence. OBJECTIVE: Our study aims to directly contribute to the existing knowledge of this novel Chinese medicine mode in older patients with cancer enrolled at the Department of Oncology/Chinese Medicine, Nanjing First Hospital, China, thereby establishing the basis for further research. METHODS: This study involves a 2-arm, prospective, randomized, assessor-blinded clinical trial in older patients with cancer experiencing chemobrain-related stress and treated with Chinese acupuncture from September 30, 2023, to December 31, 2025. We will enroll 168 older patients with cancer with clinically confirmed chemobrain. These participants will be recruited through screening by oncologists for Chinese acupuncture therapy and evaluation. Electroacupuncture will be performed by a registered practitioner of Chinese medicine. The electroacupuncture intervention will take about 30 minutes every session (2 sessions per week over 8 weeks). For the experimental group, the acupuncture points are mainly on the head, limbs, and abdomen, with a total of 6 pairs of electrically charged needles on the head, while for the control group, the acupuncture points are mainly on the head and limbs, with only 1 pair of electrically charged needles on the head. RESULTS: Eligible participants will be randomized to the control group or the experimental group in 1:1 ratio. The primary outcome of this intervention will be the scores of the Montreal Cognitive Assessment. The secondary outcomes, that is, attentional function and working memory will be determined by the Digit Span Test scores. The quality of life of the patients and multiple functional assessments will also be evaluated. These outcomes will be measured at 2, 4, 6, and 8 weeks after the randomization. CONCLUSIONS: This efficacy trial will explore whether Chinese electroacupuncture can prevent chemobrain, alleviate the related symptoms, and improve the quality of life of older patients with cancer who are undergoing or are just going to begin chemotherapy. The safety of this electroacupuncture intervention for such patients will also be evaluated. Data from this study will be used to promote electroacupuncture application in patients undergoing chemotherapy and support the design of further real-world studies. TRIAL REGISTRATION: ClinicalTrials.gov NCT05876988; https://clinicaltrials.gov/ct2/show/NCT05876988. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/53853.

Solution blow spun bilayer chitosan/polylactic acid nanofibrous patch with antibacterial and anti-inflammatory properties for accelerating acne healing.

The quercetin (QC) loaded chitosan (CS) nanofibrous patch (CSQC) was designed and fabricated successfully by solution blow spinning (SBS). And it was employed to explore a functional double-layer nanofibrous patch (CSQC/PLA) with polylactic acid (PLA) for overcoming the resistance of acne-causing bacteria to antibiotics and local cutaneous irritation. The nanofibrous patch possessed a fluffy bilayer structure with good air permeability, which may be befitted from the SBS method. The 10 % QC loaded CSQC0.10/PLA had sustained release ability of QC for 24 h. A high free radical clearance rate (91.18 ± 2.26 %) and robust antibacterial activity against P. acnes (94.4 %) were achieved for CSQC0.10/PLA with excellent biocompatibility. Meanwhile, E. coli and S. aureus were also suppressed with 99.4 % and 99.2 %, respectively. Moreover, the expression of pro-inflammatory cytokines (IL-6 and TNF-α) was significantly reduced, conducive to acne healing. Therefore, the CSQC0.10/PLA bilayer nanofibrous patch designed here may shed some light on developing multifunctional materials for treating acne infectious wounds.

ChatGPT Performs on the Chinese National Medical Licensing Examination.

ChatGPT, a language model developed by OpenAI, uses a 175 billion parameter Transformer architecture for natural language processing tasks. This study aimed to compare the knowledge and interpretation ability of ChatGPT with those of medical students in China by administering the Chinese National Medical Licensing Examination (NMLE) to both ChatGPT and medical students. We evaluated the performance of ChatGPT in three years' worth of the NMLE, which consists of four units. At the same time, the exam results were compared to those of medical students who had studied for five years at medical colleges. ChatGPT's performance was lower than that of the medical students, and ChatGPT's correct answer rate was related to the year in which the exam questions were released. ChatGPT's knowledge and interpretation ability for the NMLE were not yet comparable to those of medical students in China. It is probable that these abilities will improve through deep learning.

The formation process and mechanism of gastrointestinal anastomosis after retrieval anchor-assisted endoscopic ultrasound-guided gastroenterostomy: a preclinical study.

BACKGROUND AND AIMS: Endoscopic ultrasound-guided gastroenterostomy (EUS-GE) is a promising method of treating gastric outlet obstruction. However, no study has investigated gastrointestinal anastomosis formation after EUS-GE. We aimed to evaluate the formation of gastrointestinal anastomosis after EUS-GE in a porcine model. METHODS: Retrieval anchor-assisted EUS-GE was performed on 15 Bama mini pigs. Five pigs each were randomly euthanized 3, 7, and 14 days postoperatively to evaluate the formation of gastrointestinal anastomosis and measure the anastomotic distance. The expression of transforming growth factor-ß1(TGF-ß1) and Smad3 in the anastomosis site were examined by immunohistochemistry. RESULTS: EUS-GE was successfully performed in all 15 pigs. The mean procedure time was 29.2 ± 6.0 (range 18-40) minutes. The anastomotic distance was 34.0 ± 3.6 cm in 14 pigs. The site of gastroenterostomy of one pig was at the ileum. For pigs euthanized 3 days postoperatively, the structure was mechanically maintained by a stent. For pigs euthanized 7 or 14 days postoperatively, the stomach and small intestine were anastomosed to form a stable structure. The level of TGF-ß1 and Smad3 in the anastomosis site gradually increased from 3 to 14 days after EUS-GE. TGF-ß1 and Smad3 expression had a significant difference between 3 days, 7 days, and 14 days after EUS-GE (P < 0.05). CONCLUSIONS: For EUS-GE, the stomach and small intestine were initially linked together mechanically and spontaneously anastomosed to form a stable structure 7 days postoperatively. TGF-ß1 and Smad3 play an important role in the formation of a stable structure of gastrointestinal anastomosis.

An ultrasensitive NIR-IIa' fluorescence-based multiplex immunochromatographic strip test platform for antibiotic residues detection in milk samples.

INTRODUCTION: Widely used in livestock breeding, residues of antibiotic drugs in milk have become a threat to food safety and human health. Current rapid detection technologies using colorimetric immunochromatographic strip tests (IST) lack the necessary sensitivity for on-site trace monitoring. Fluorescence-based detection in the near-infrared IIa' (NIR-IIa') region (1000 âˆ¼ 1300 nm) has enormous potential due to greatly minimized auto-fluorescence and light scattering. OBJECTIVES: The aim of this work is to develop an ultrasensitive IST platform using NIR-IIa' fluorescent nanoparticles as labels for multiplex antibiotic residues detection in milk. METHODS: NIR-IIa' fluorescent nanoparticles were assembled by encapsulating synthesized NIR-IIa' fluorophores into carboxyl - modified polystyrene nanoparticles. The NIR-IIa' nanoparticles were subsequently used as labels in an IST platform to detect sulfonamides, quinolones, and lincomycin simultaneously in milk. A portable fluorescent reader was fabricated to provide on-site detection. To further validate the developed IST platform, the detection was compared with LC-MS/MS in 22 real milk samples. RESULTS: Fluorescent nanoparticles were synthesized with low energy emission (1030 nm) and large Stokes shift (>250 nm) showing a much higher signal-to-noise ratio compared with fluorophores emitting in the NIR-I region. The developed IST platform yielded a highly sensitive, simultaneous quantification of sulfonamides, quinolones, and lincomycin in milk with detection limits of 46.7, 27.6 and 51.4 pg/mL, respectively, achieving a wide detection range (up to 50 ng/mL). The IST platform showed good accuracy, reproducibility, and specificity with the portable fluorescent reader which could rapidly quantify in 10 s. These results were better than reported immunochromatographic assays using fluorescent labels, and remarkably, showed a higher recognition ability than LC-MS/MS for real samples. CONCLUSION: The utility of NIR-IIa' fluorescence-based IST platform for the fast, sensitive, and accurate detection of antibiotics in milk was demonstrated, successfully verifying the potential of this platform in detecting trace materials in complex matrices.

Breaking the Electronic Conductivity Bottleneck of Manganese Oxide Family for High-Power Fluorinated Graphite Composite Cathode by Ligand-Field High-Dimensional Constraining Strategy.

Primary lithium fluorinated graphite (Li/CFx ) batteries with superior energy density are an indispensable energy supply for multiple fields but suffer from sluggish reaction kinetics of the CFx cathode. Designing composite cathodes emerges as a solution to this problem. Despite the optimal composite component for CFx , the manganese oxide family represented by MnO2 is still faced with an intrinsic electronic conductivity bottleneck, which severely limits the power density of the composite cathode. Here, a cation-induced high-dimensional constraining strategy from the perspective of ligand-field stacking structure topological design, which breaks the molecular orbital hybridization of pristine semiconductive oxides to transform them into the high-conductivity metallic state while competitively maintaining structural stability, is proposed. Through first-principles phase diagram calculations, mixed-valent Mn5 O8 ( Mn 2 2 + Mn 3 4 + O 8 ${\rm{Mn}}_2^{2 + }{\rm{Mn}}_3^{4 + }{{\rm{O}}_8}$ ) is explored as an ideal high-dimensional constraining material with satisfied conductivity and large-scale production feasibility. Experiments demonstrate that the as-proposed CFx  @ Mn5 O8 composite cathode achieves 2.36 times the power density (11399 W kg-1 ) of pristine CFx and a higher CFx conversion ratio (86%). Such a high-dimensional field-constraining strategy is rooted in the established four-quadrant electronic structure tuning framework, which fundamentally changes the orbital symmetry under the ligand field to overcome the common conductivity challenge of wide transition metal oxide materials.