Me-LLaMA: Medical Foundation Large Language Models for Comprehensive Text Analysis and Beyond.

Recent advancements in large language models (LLMs) like ChatGPT and LLaMA have shown significant potential in medical applications, but their effectiveness is limited by a lack of specialized medical knowledge due to general-domain training. In this study, we developed Me-LLaMA, a new family of open-source medical LLMs that uniquely integrate extensive domain-specific knowledge with robust instruction-following capabilities. Me-LLaMA comprises foundation models (Me-LLaMA 13B and 70B) and their chat-enhanced versions, developed through comprehensive continual pretraining and instruction tuning of LLaMA2 models using both biomedical literature and clinical notes.

Membrane-free Electrolysis Production of Hydrogen Peroxide on Low-cost Metal-free Electrocatalysts for Dye Degradation.

The efficient production of hydrogen peroxide (HO) solution was achieved by combining cathodic two-electron oxygen reduction (2e ORR) and anodic two-electron water oxidation (2e WOR) in two half-reaction cells. h-BN loaded on carbon fibers (h-BN@C) is prepared and employed as an anode material to catalyze 2e WOR, while sulfonated commercial BP-2000 carbons (BP-2000-SOH) were prepared as the cathode materials for 2e ORR. In a 2 M KHCO solution, an overall Faradaic efficiency of 97 % and a total HO production rate of 1872 mmol g h over metal-free electrodes were accomplished in a membrane-free flow cell.

Changing Benzoxazole Ring into Nonring Imine Linkages on Covalent Organic Frameworks with Tuning HO Photosynthesis Performance.

Two π-conjugated covalent organic frameworks (COFs) with nonring imine or benzoxazole ring linkages were prepared by reacting 3,3'-dihydrooxybenzidine (BDOH) with 3,5-triformylbenzene (Tb) in the presence or absence of benzimidazole (BDOH-Tb- and BDOH-Tb-). Although two COFs indicated similar composition, crystalline structures, and morphologies, imine-based BDOH-Tb- exhibited a photocatalytic HO production rate of 2490 μmol·g·h in sacrificial reagent-free pure water, higher than that of benzoxazole-based BDOH-Tb- (1168 μmol·g·h). The higher photocatalytic activity of BDOH-Tb- was attributed to its more efficient photoinduced charge separation and utilization efficiency and different 2e ORR active sites over the two COFs.

A Study of Biomedical Relation Extraction Using GPT Models.

Relation Extraction (RE) is a natural language processing (NLP) task for extracting semantic relations between biomedical entities. Recent developments in pre-trained large language models (LLM) motivated NLP researchers to use them for various NLP tasks. We investigated GPT-3.

Me-LLaMA: Foundation Large Language Models for Medical Applications.

Recent advancements in large language models (LLMs) such as ChatGPT and LLaMA have hinted at their potential to revolutionize medical applications, yet their application in clinical settings often reveals limitations due to a lack of specialized training on medical-specific data. In response to this challenge, this study introduces Me-LLaMA, a novel medical LLM family that includes foundation models - Me-LLaMA 13/70B, along with their chat-enhanced versions - Me-LLaMA 13/70B-chat, developed through continual pre-training and instruction tuning of LLaMA2 using large medical datasets. Our methodology leverages a comprehensive domain-specific data suite, including a large-scale, continual pre-training dataset with 129B tokens, an instruction tuning dataset with 214k samples, and a new medical evaluation benchmark (MIBE) across six critical medical tasks with 12 datasets.

Advancing entity recognition in biomedicine via instruction tuning of large language models.

Motivation: Large Language Models (LLMs) have the potential to revolutionize the field of Natural Language Processing, excelling not only in text generation and reasoning tasks but also in their ability for zero/few-shot learning, swiftly adapting to new tasks with minimal fine-tuning. LLMs have also demonstrated great promise in biomedical and healthcare applications. However, when it comes to Named Entity Recognition (NER), particularly within the biomedical domain, LLMs fall short of the effectiveness exhibited by fine-tuned domain-specific models.

mNFE: microbiome network flow entropy for detecting pre-disease states of type 1 diabetes.

In the development of Type 1 diabetes (T1D), there are critical states just before drastic changes, and identifying these pre-disease states may predict T1D or provide crucial early-warning signals. Unlike gene expression data, gut microbiome data can be collected noninvasively from stool samples. Gut microbiome sequencing data contain different levels of phylogenetic information that can be utilized to detect the tipping point or critical state in a reliable manner, thereby providing accurate and effective early-warning signals.

Improving large language models for clinical named entity recognition via prompt engineering.

Importance: The study highlights the potential of large language models, specifically GPT-3.5 and GPT-4, in processing complex clinical data and extracting meaningful information with minimal training data. By developing and refining prompt-based strategies, we can significantly enhance the models' performance, making them viable tools for clinical NER tasks and possibly reducing the reliance on extensive annotated datasets.

MIWE: detecting the critical states of complex biological systems by the mutual information weighted entropy.

Complex biological systems often undergo sudden qualitative changes during their dynamic evolution. These critical transitions are typically characterized by a catastrophic progression of the system. Identifying the critical point is critical to uncovering the underlying mechanisms of complex biological systems.

Specific network information gain for detecting the critical state of colorectal cancer based on gut microbiome.

There generally exists a critical state or tipping point from a stable state to another in the development of colorectal cancer (CRC) beyond which a significant qualitative transition occurs. Gut microbiome sequencing data can be collected non-invasively from fecal samples, making it more convenient to obtain. Furthermore, intestinal microbiome sequencing data contain phylogenetic information at various levels, which can be used to reliably identify critical states, thereby providing early warning signals more accurately and effectively.

Information-Motivation-Behavioral Skills Model Supplemented With the Moderated-Mediation Path: A Framework for Interpreting Patients' Online Medical Services Utilization.

Purpose: This study aimed to explore the internal determinants affecting patients' utilization of online medical services (OMS) based on the information-motivation-behavioral skills model from a behavioral perspective.

Design: A cross-sectional study.

Setting: This study was conducted in three medical institutions in Jiangsu Province, China.

Behavioral Interpretation of Willingness to Use Wearable Health Devices in Community Residents: A Cross-Sectional Study.

Wearable health devices (WHDs) have become increasingly advantageous in long-term health monitoring and patient management. However, most people have not yet benefited from such innovative technologies, and the willingness to accept WHDs and their influencing factors are still unclear. Based on two behavioral theories: the theory of planned behavior (TPB) and the diffusion of innovation (DOI), this study aims to explore the influencing factors of willingness to use WHDs in community residents from the perspective of both internal and external factors.

Weakly Hydrophilic Imine-Linked Covalent Benzene-Acetylene Frameworks for Photocatalytic HO Production in the Two-Phase System.

Conversing oxygen (O) to hydrogen peroxide (HO) driven by solar energy is a promising HO onsite production route but often short of efficient and durable photocatalysts. Herein, strong π-π conjugate polycyclic aromatic benzene and acetylene units have been constructed into new covalent organic frameworks (COFs) linked by imine C═N bonding. These COFs demonstrated two-dimensional hexagonal crystalline frameworks with higher crystallinity and larger surface area (>600 m g).

Empirical validation of the information-motivation-behavioral skills model of gestational weight management behavior: a framework for intervention.

Background: Unhealthy gestational weight gain is a modifiable risk factor for adverse maternal and child health. Appropriate and effective intervention strategies that focus on behavioral change or maintenance are critical in weight management during pregnancy. Our aim was to uncover the influencing factors and psychosocial mechanisms of gestational weight control behavior, and to construct a behavioral model suitable for intervention based on Information-Motivation-Behavioral skills (IMB) model.

Behavioral Intention and Its Predictors toward COVID-19 Booster Vaccination among Chinese Parents: Applying Two Behavioral Theories.

The booster vaccination of COVID-19 is being implemented in most parts of the world. This study used behavioral psychology to investigate the predictors of parents' intentions regarding the COVID-19 booster vaccination for their children. This is a cross-sectional study with a self-designed questionnaire based on two behavioral theories-protective motivation theory (PMT) and theory of planned behavior (TPB).

Determinants of excessive gestational weight gain: a systematic review and meta-analysis.

Background: The prevalence of excessive gestational weight gain (EGWG) during pregnancy is increasing, and it is extremely harmful to pregnant women and newborns. Previous studies have suggested that EGWG is associated with various factors. We conducted a systematic review and meta-analysis to identify, quantify and analyze determinants of EGWG and evaluate the effect of these determinants on EGWG.

Cognitive Factors of Weight Management During Pregnancy Among Chinese Women: A Study Applying Protective Motivation Theory.

Purpose: This study aimed to explore the psychological cognitive factors of weight management during pregnancy based on protective motivation theory (PMT).

Design: Cross-sectional study.

Setting: Participants were recruited at the Maternal and Child Health Hospital of Changzhou City, Jiangsu Province, China.

Analysis of the Weight Management Behavior of Chinese Pregnant Women: An Integration of the Protection Motivation Theory and the Information-Motivation-Behavioral Skills Model.

Inappropriate gestational weight gain has become a public health concern that threatens maternal and child health. Pregnant women's ability to manage their weight during pregnancy directly impacts their weight gain. In this study, we integrated the protection motivation theory and the information-motivation-behavioral skills model to develop an integrative theoretical model suitable for pregnancy weight management and reveal significant explainable factors of weight management behaviors during pregnancy.

Behaviour model integrated by protection motivation theory and information-motivation-behavioural skills model applying in pregnancy weight management (PrInMAMa): a study protocol for a randomised controlled trial in China.

Introduction: Excessive gestational weight gain poses a significant threat to maternal and child health. The healthy behaviour theory has been increasingly applied to weight management during pregnancy, but research is still insufficient. The successful application of the protection motivation theory (PMT) and the information-motivation-behavioural skills (IMB) model in the field of healthy behaviour laid the foundation for this intervention study.

Assessing Determinants of Online Medical Services Adoption Willingness of General Hospital Physicians Using the Information-Motivation-Behavioral Skills Model: A Multi-Group Structural Equation Modeling Approach.

Purpose: Physician adoption of online medical services (OMS) has been hastened by the COVID-19 pandemic, but their adoption willingness still requires to be improved. This study aims to construct a physician's OMS adoption willingness model based on the information-motivation-behavioral skill (IMB) theory, explore the determinants affecting adoption willingness and its influencing pathways, and evaluate the moderating effects of OMS use experience on willingness through multi-group analysis.

Participants And Methods: A cross-sectional survey was conducted among physicians in three public hospitals of Jiangsu province, China, from June to July 2020, using a multi-stage sampling method.

Attention-Based Deep Multiple-Instance Learning for Classifying Circular RNA and Other Long Non-Coding RNA.

Circular RNA (circRNA) is a distinguishable circular formed long non-coding RNA (lncRNA), which has specific roles in transcriptional regulation, multiple biological processes. The identification of circRNA from other lncRNA is necessary for relevant research. In this study, we designed attention-based multi-instance learning (MIL) network architecture fed with a raw sequence, to learn the sparse features of RNA sequences and to accomplish the circRNAs identification task.

Repeatability and Reproducibility of Computed Tomography Radiomics for Pulmonary Nodules: A Multicenter Phantom Study.

Background: Radiomics can yield minable information from medical images, which can facilitate computer-aided diagnosis. However, the lack of repeatability and reproducibility of radiomic features (RFs) may hinder their generalizability in clinical applications.

Objectives: The aims of this study were to explore 3 main sources of variability in RFs, investigate their influencing magnitudes and patterns, and identify a subset of robust RFs for further studies.