Benefiting from rich knowledge and the exceptional ability to understand text, large language models like ChatGPT have shown great potential in English clinical environments. However, performance of non-English settings, as well its reasoning, not been explored depth.

Large Language Models (LLMs) such as ChatGPT and Med-PaLM have excelled in various medical question-answering tasks. However, these English-centric models encounter challenges non-English clinical settings, primarily due to limited knowledge respective languages, a consequence of imbalanced training corpora. We systematically evaluate LLMs the Chinese context develop novel in-context learning framework enhance their performance.

Social media-based public health research is crucial for epidemic surveillance, but most studies identify relevant corpora with keyword-matching. This study develops a system to streamline the process of curating colloquial medical dictionaries. We demonstrate pipeline by Unified Medical Language System (UMLS)-colloquial symptom dictionary from COVID-19-related tweets as proof concept.

For an emergent pandemic, such as COVID-19, the statistics of symptoms based on hospital data may be biased or delayed due to high proportion asymptomatic mild-symptom infections that are not recorded in hospitals. Meanwhile, difficulty accessing large-scale clinical also limits many researchers from conducting timely research.Given wide coverage and promptness social media, this study aimed present efficient workflow track visualize dynamic characteristics co-occurrence for COVID-19...

The COVID-19 pandemic has caused substantial damage to global health. Even though three years have passed, the world continues struggle with virus. Concerns are growing about impact of on mental health infected individuals, who more likely experience depression, which can long-lasting consequences for both affected individuals and world. Detection intervention at an early stage reduce risk depression in patients. In this paper, we investigated relationship between infection through social...

Large pre-trained models have revolutionized natural language processing (NLP) research and applications, but high training costs limited data resources prevented their benefits from being shared equally amongst speakers of all the world's languages. To address issues cross-linguistic access to such reduce energy consumption for sustainability during large-scale model training, this study proposes an effective energy-efficient framework called GreenPLM that uses bilingual lexicons directly...

In this paper, we propose a multi-input multi-output beamforming transmit optimization model for joint radar sensing and multi-user communications, where the design of beamformers is formulated as an problem whose objective weighted combination sum rate Cramér-Rao bound, subject to power budget constraint. Obtaining global solution challenging task, because maximization itself (even without considering metric) known be NP-hard. efficient branch-and-bound algorithm solving based on McCormick...

Clinical reasoning refers to the cognitive process that physicians employ in evaluating and managing patients. This typically involves suggesting necessary examinations, diagnosing patients’ diseases, selecting appropriate therapies, etc. Accurate clinical requires extensive medical knowledge rich experience, setting a high bar for physicians. is particularly challenging developing countries due overwhelming number of patients limited physician resources, contributing significantly global...

In this paper, we consider the joint multicast and unicast beamforming design problem in multi-input single-output downlink wireless network, where all base stations (BSs) potentially can cooperate (as a single virtual BS) to transmit common multiple dedicated data streams at same time on frequency band. To reduce cooperation overhead (among different BSs), prefer partial transmission such that each user's stream (either or unicast) is served by only small subset of BSs. We formulate from...

Nucleic acid testing (NAT) has been widely used in many fields such as medical diagnosis, food safety and forensic identification. However, it can only be carried out professional laboratory because the test process is complicated rigorous. In this paper, a nucleic amplification system based on polymerase chain reaction (PCR) was developed to meet requirements of point-of-care (POCT) for acids. Firstly, mechanical structure electronic control were designed constructed. Secondly, an integral...

Quantized constant envelope (QCE) transmission is a popular and effective technique to reduce the hardware cost improve power efficiency of 5G beyond systems equipped with large antenna arrays. It has been widely observed that number quantization levels substantial impact on system performance. This paper aims quantify Specifically, we consider downlink single-user multiple-input-single-output (MISO) <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML"...

With appropriate data selection and training techniques, Large Language Models (LLMs) have demonstrated exceptional success in various medical examinations multiple-choice questions. However, the application of LLMs dialogue generation-a task more closely aligned with actual practice-has been less explored. This gap is attributed to insufficient knowledge LLMs, which leads inaccuracies hallucinated information generated responses. In this work, we introduce Medical Knowledge enhancement...

Abstract Objective To evaluate the effectiveness and reasoning ability of ChatGPT in diagnosing retinal vascular diseases Chinese clinical environment. Materials Methods We collected 1226 fundus fluorescein angiography reports corresponding diagnosis written Chinese, tested with four prompting strategies (direct or explanation English). Results using English prompt for direct achieved best performance, F1-score 80.05%, which was inferior to ophthalmologists (89.35%) but close ophthalmologist...

We introduce YATO, an open-source, easy-to-use toolkit for text analysis with deep learning. Different from existing heavily engineered toolkits and platforms, YATO is lightweight user-friendly researchers cross-disciplinary areas. Designed in a hierarchical structure, supports free combinations of three types widely used features including 1) traditional neural networks (CNN, RNN, etc.); 2) pre-trained language models (BERT, RoBERTa, ELECTRA, 3) user-customized via simple configurable file....

The COVID-19 pandemic continues to bring up various topics discussed or debated on social media. In order explore the impact of pandemics people's lives, it is crucial understand public's concerns and attitudes towards pandemic-related entities (e.g., drugs, vaccines) However, models trained existing named entity recognition (NER) targeted sentiment analysis (TSA) datasets have limited ability COVID-19-related media texts because these are not designed annotated from a medical perspective....

$\textbf{Objectives}$: Large Language Models (LLMs) such as ChatGPT and Med-PaLM have excelled in various medical question-answering tasks. However, these English-centric models encounter challenges non-English clinical settings, primarily due to limited knowledge respective languages, a consequence of imbalanced training corpora. We systematically evaluate LLMs the Chinese context develop novel in-context learning framework enhance their performance. $\textbf{Materials Methods}$: The latest...

In this paper, we propose a multi-input multi-output transmit beamforming optimization model for joint radar sensing and multi-user communications, where the design of beamformers is formulated as an problem whose objective weighted combination sum rate Cram\'{e}r-Rao bound, subject to power budget. Obtaining global solution nonconvex challenging task, since sum-rate maximization itself (even without considering metric) known be NP-hard. The main contributions paper are threefold. Firstly,...

Accurate and transparent financial information disclosure is crucial in the fields of accounting finance, ensuring market efficiency investor confidence. Among many platforms, Chinese stock exchanges' interactive platform provides a novel way for listed firms to disclose interest investors through an online question-and-answer (Q&A) format. However, it common respond questions with limited or no substantive information, automatically evaluating quality on large amounts Q&A pairs challenging....

Social media is recognized as an important source for deriving insights into public opinion dynamics and social impacts due to the vast textual data generated daily 'unconstrained' behavior of people interacting on these platforms. However, such analyses prove challenging semantic shift phenomenon, where word meanings evolve over time. This paper proposes unsupervised dynamic embedding method capture longitudinal shifts in without predefined anchor words. The leverages co-occurrence...

In this paper, we consider a class of convex programming problems with linear equality constraints, which finds broad applications in machine learning and signal processing. We propose new adaptive balanced augmented Lagrangian (ABAL) method for solving these problems. The proposed ABAL adaptively selects the stepsize parameter enjoys low per-iteration complexity, involving only computation proximal mapping objective function solution equation. These features make well-suited to large-scale...