检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Chen, Xingran Wu, Zhenke Shi, Xu Cho, Hyunghoon Mukherjee, Bhramar Department of Biostatistics University of Michigan United States Department of Biomedical Informatics and Data Science Yale University United States Department of Biostatistics Yale University United States

Objectives: To conduct a scoping review of existing approaches for synthetic Electronic Health Records (EHR) data generation, to benchmark major methods with proposed open-source software, and to offer recommendations for practitioners. Materials and Methods: We search three academic databases for our scoping review. Methods are benchmarked on open-source EHR datasets, Medical Information Mart for Intensive Care III and IV (MIMIC-III/IV). Seven existing methods covering major categories and two baseline methods are implemented and compared. Evaluation metrics concern data fidelity, downstream utility, privacy protection, and computational cost. Results: 42 studies are identified and classified into five categories. Seven open-source methods covering all categories are selected, trained on MIMIC-III, and evaluated on MIMIC-III or MIMIC-IV for transportability considerations. Among them, Generative Adversarial Network (GAN)-based methods demonstrate competitive performance in fidelity and utility on MIMIC-III;rule-based methods excel in privacy protection. Similar findings are observed on MIMIC-IV, except that GAN-based methods further outperform the baseline methods in preserving fidelity. A Python package, "SynthEHRella", is provided to integrate various choices of approaches and evaluation metrics, enabling more streamlined exploration and evaluation of multiple methods, which can be accessed at https://***/chenxran/synthEHRella. Discussion: Method choice is governed by the relative importance of the evaluation metrics in downstream use cases. We provide a decision tree to guide the choice among the benchmarked methods. Conclusion: GAN-based methods excel when distributional shifts exist between the training and testing populations. Otherwise, CorGAN and MedGAN are most suitable for association modeling and predictive modeling, respectively. Future research should prioritize enhancing fidelity of the synthetic data while controlling privacy exposure, and com

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

LITA: An Efficient LLM-assisted Iterative Topic Augmentation Framework

arXiv

引用

arXiv 2024年

作者： Chang, Chia-Hsuan Tsai, Jui-Tse Tsai, Yi-Hang Hwang, San-Yih Department of Biomedical Informatics & Data Science Yale University New Haven United States Department of Information Management National Sun Yat-sen University Kaohsiung Taiwan

Topic modeling is widely used for uncovering thematic structures within text corpora, yet traditional models often struggle with specificity and coherence in domain-focused applications. Guided approaches, such as SeededLDA and CorEx, incorporate user-provided seed words to improve relevance but remain labor-intensive and static. Large language models (LLMs) offer potential for dynamic topic refinement and discovery, yet their application often incurs high API costs. To address these challenges, we propose the LLM-assisted Iterative Topic Augmentation framework (LITA), an LLM-assisted approach that integrates user-provided seeds with embedding-based clustering and iterative refinement. LITA identifies a small number of ambiguous documents and employs an LLM to reassign them to existing or new topics, minimizing API costs while enhancing topic quality. Experiments on two datasets across topic quality and clustering performance metrics demonstrate that LITA outperforms five baseline models, including LDA, SeededLDA, CorEx, BERTopic, and PromptTopic. Our work offers an efficient and adaptable framework for advancing topic modeling and text clustering. © 2024, CC BY.

关键词： Modeling languages

来源：评论

学校读者我要写书评

暂无评论

Higher-Order Newton Methods with Polynomial Work per Iteration

arXiv

引用

arXiv 2023年

作者： Ahmadi, Amir Ali Chaudhry, Abraar Zhang, Jeffrey Princeton University Operations Research and Financial Engineering United States Yale University Department of Biomedical Informatics and Data Science United States

We present generalizations of Newton’s method that incorporate derivatives of an arbitrary order d but maintain a polynomial dependence on dimension in their cost per iteration. At each step, our dth-order method uses semidefinite programming to construct and minimize a sum of squares-convex approximation to the dth-order Taylor expansion of the function we wish to minimize. We prove that our dth-order method has local convergence of order d. This results in lower oracle complexity compared to the classical Newton method. We show on numerical examples that basins of attraction around local minima can get larger as d increases. Under additional assumptions, we present a modified algorithm, again with polynomial cost per iteration, which is globally convergent and has local convergence of order d. Copyright © 2023, The Authors. All rights reserved.

关键词： Newton-Raphson method

来源：评论

学校读者我要写书评

暂无评论

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

arXiv

引用

arXiv 2024年

作者： Xie, Qianqian Chen, Qingyu Chen, Aokun Peng, Cheng Hu, Yan Lin, Fongci Peng, Xueqing Huang, Jimin Zhang, Jeffrey Keloth, Vipina Zhou, Xinyu Qian, Lingfei He, Huan Shung, Dennis Ohno-Machado, Lucila Wu, Yonghui Xu, Hua Bian, Jiang Department of Biomedical Informatics and Data Science Yale School of Medicine Yale University New HavenCT United States Department of Health Outcomes and Biomedical Informatics College of Medicine University of Florida GainesvilleFL United States School of Biomedical Informatics University of Texas Health Science Center at Houston HoustonTX United States Yale School of Medicine Yale University New HavenCT United States

Background Recent advancements in large language models (LLMs) like ChatGPT and LLaMA have shown promise in medical applications, though their performance in medical language understanding still requires enhancement. This study aims to develop foundational medical LLMs by training open-source LLaMA models with large-scale, domain-specific datasets to enhance their efficacy across a variety of medical text analysis tasks and medical diagnosis. Methods We developed Me-LLaMA, a new medical LLM family that includes foundation models - Me-LLaMA 13/70B, and their chat-enhanced versions, through continual pre-training and instruction tuning of LLaMA2 using both biomedical literature and clinical notes. Me-LLaMA utilized the largest and most comprehensive medical data, including 129B pre-training tokens and 214K instruction tuning samples from diverse biomedical and clinical data sources. Training the 70B models required substantial computational resources, exceeding 100,000 A100 GPU hours. We applied Me-LLaMA to six medical text analysis tasks and evaluated its performance on 12 benchmark datasets. To further assess Me-LLaMA's potential clinical utility, we evaluated its performance on complex clinical case diagnosis compared with other commercial LLMs, using both automatic and human evaluations. Results Me-LLaMA models outperform LLaMA, and other existing open-source medical LLMs in both zero-shot and supervised learning settings for most text analysis tasks. With task-specific instruction tuning, Me-LLaMA models also surpass leading commercial LLMs, outperforming ChatGPT on 7 out of 8 datasets and GPT-4 on 5 out of 8 datasets. Moreover, for diagnosing complex clinical cases, Me-LLaMA's performance is comparable to ChatGPT and GPT-4. Conclusion Domain-specific data is crucial for building medical foundation LLMs that enhance diverse downstream text analysis tasks and medical applications. The substantial computing costs associated with training such models require careful

关键词： Medical computing

来源：评论

学校读者我要写书评

暂无评论

Suicide Phenotyping from Clinical Notes in Safety-Net Psychiatric Hospital Using Multi-Label Classification with Pre-Trained Language Models

arXiv

引用

arXiv 2024年

作者： Li, Zehan Hu, Yan Lane, Scott Selek, Salih Shahani, Lokesh Machado-Vieira, Rodrigo Soares, Jair Xu, Hua Liu, Hongfang Huang, Ming MacWilliam School of Biomedical Informatics The University of Texas Health Science Center HoustonTX United States Department of Psychiatry & Behavioral Sciences McGovern Medical School The University of Texas Health Science at Houston HoustonTX United States Section of Biomedical Informatics and Data Science School of Medicine Yale University New HavenCT United States

Accurate identification and categorization of suicidal events can yield better suicide precautions, reducing operational burden, and improving care quality in high-acuity psychiatric settings. Pre-trained language models offer promise for identifying suicidality from unstructured clinical narratives. We evaluated the performance of four BERT-based models using two fine-tuning strategies (multiple single-label and single multi-label) for detecting coexisting suicidal events from 500 annotated psychiatric evaluation notes. The notes were labeled for suicidal ideation (SI), suicide attempts (SA), exposure to suicide (ES), and non-suicidal self-injury (NSSI). RoBERTa outperformed other models using binary relevance (acc=0.86, F1=0.78). MentalBERT (F1=0.74) also exceeded BioClinicalBERT (F1=0.72). RoBERTa fine-tuned with a single multi-label classifier further improved performance (acc=0.88, F1=0.81), highlighting that models pre-trained on domain-relevant data and the single multi-label classification strategy enhance efficiency and performance. © 2024, CC BY.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

CFGNet: Coarse-to-Fine Grained Network for Image Classification of Confusing Fly Species 7

CFGNet: Coarse-to-Fine Grained Network for Image Classificat...

引用

7th International Conference on Pattern Recognition and Artificial Intelligence, PRAI 2024

作者： Niu, Lin Wang, Bo Niu, Baomin Zhang, Jin Liang, Danning Wu, Jingpeng Zhou, Rong Deng, Jianqiang School of Biomedical Informatics and Engineering Hainan Medical University Haikou China Hainan Vocational College of Political Science and Law Hainan Gongping Forensic Center Haikou China School of Big Data and Artificial Intelligence Zhengzhou University of Science and Technology Zhengzhou China Center for Educational Technology Hainan Medical University Haikou China Medicine of Hainan Medical University Department of Forensic Haikou China

ISBN: (纸本)9798350350890

In the realm of fine-grained image classification (FGIC), despite substantial technological advancements, models frequently misclassify certain fly species with exceedingly high inter-class similarity, thereby impacting overall classification accuracy. To address this challenge, we propose an innovative Coarse-to-Fine Grained Network Framework (CFGNet). This architecture tackles the issue through two stages: Initially, to alleviate classification complexity, closely related and easily confused categories are amalgamated. Subsequently, a lightweight coarse-grained information extraction network (CM) is introduced to enhance the classification accuracy of samples with relatively larger class differences. In the second stage, for precise identification of challenging samples, a fine-grained feature extraction network (FM) is proposed exclusively for difficult-to-distinguish images. Additionally, we introduce a novel feature reuse method to deeply mine and repurpose coarse-grained features from the previous stage to guide fine-grained classification and improve performance in this stage. To validate our approach, we curated a dataset based on merged confusion classes annotation (Flies-MCC). Experimental results demonstrate that our proposed network architecture achieves significant advancements in fine-grained fly classification and offers a pioneering solution for precise fly identification. © 2024 IEEE.

关键词： Coarse-grained modeling

来源：评论

学校读者我要写书评

暂无评论

Development of a Natural Language Processing Tool to Extract Acupuncture Point Location Terms

Development of a Natural Language Processing Tool to Extract...

引用

IEEE International Conference on Healthcare informatics (ICHI)

作者： Yiming Li Xueqing Peng Jianfu Li Suyuan Peng Donghong Pei Cui Tao Hua Xu Na Hong School of Biomedical Informatics The University of Texas Health Science Center at Houston Houston TX USA Section of Biomedical Informatics and Data Science School of Medicine Yale University New Haven CT USA National Institute of Health Data Science Peking University Beijing China The University of Texas MD Anderson Cancer Center Houston TX USA

Acupuncture point location information is significant to the effect of acupuncture therapy. A wealth of knowledge related to acupuncture point locations is typically dispersed across books or reports in free-text format, which can pose challenges for machines to understand and facilitate. In this study, our purpose is to develop a natural language processing (NLP) tool to extract acupuncture point location information. We adopted the World Health Organization (WHO) Standard Acupuncture Point Locations in the Western Pacific Region (WHO Standard) as our reference standard and corpus. Our proposed approach to develop and validate the NLP tool involves three key steps: 1) We leveraged pre-existing terminologies from the anatomy domain, specifically the International Anatomical Terminology (IAT) and Uber-anatomy Ontology (UBERON), to construct a baseline dictionary of acupoint locations. This dictionary was then used in conjunction with a dictionary lookup method to match terms in the acupuncture location descriptions outlined by the WHO Standard; 2) We employed NLP models based on machine learning and deep learning techniques, specifically utilizing Conditional Random Fields (CRFs) and Recurrent Neural Network (RNN) for training. The CRFs and RNN models achieved overall F1 scores of 0.955 and 0.945, respectively; 3) We encoded the extracted entities into standardized concepts and enriched the baseline dictionary using these entities. In this study, a customized NLP tool was developed and validated for extracting acupuncture point location terms. This tool can be used in various informatics applications to support acupuncture research, education, and practices.

关键词：

来源：评论

学校读者我要写书评

暂无评论

EMPATHIC: Emulating Human-Like Multimodal Personality Architecture Through Thoughtful Human-AI Conversation

EMPATHIC: Emulating Human-Like Multimodal Personality Archit...

引用

International Conference on Confluence The Next Generation Information Technology Summit (Confluence)

作者： Varshini Devi I I. R. Oviya Kalpana Raja Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Chennai India Section for Biomedical Informatics and Data Science School of Medicine Yale University New Haven Connecticut USA

In the evolving landscape of human-computer interaction, this paper introduces an innovative framework poised to revolutionize chatbot systems. Our framework, meticulously designed for emotionally aware multimodal chatbots, seamlessly integrates real-time face and emotion recognition with natural language processing. Currently in the developmental phase, the framework exhibits a distinctive proficiency in executing a diverse array of human instructions. These instructions span tasks such as generating detailed captions, object counting, and responding to general queries. Leveraging parameter-efficient fine-tuning from OpenFlamingo, augmented by the inclusion of the Low-rank Adapter (LoRA), our framework strategically prioritizes versatility and operational efficiency. The core of our approach lies in the establishment of instruction templates that intricately merge both vision and language data. This strategic integration forms the backbone of our multi-modality instruction tuning methodology, enhancing the adaptability of the chatbot. Recognizing the pivotal role of high-quality training data, we emphasize its significance in shaping the dialogue performance of the chatbot. As the framework progresses through its developmental stages, this research lays a robust foundation for the creation of a versatile conversational agent. The envisioned applications of this agent range from enhancing customer service experiences to providing valuable support in mental health scenarios.

关键词：

来源：评论

学校读者我要写书评

暂无评论

CEHR-GPT: Generating Electronic Health Records with Chronological Patient Timelines

arXiv

引用

arXiv 2024年

作者： Pang, Chao Jiang, Xinzhuo Pavinkurve, Nishanth Parameshwar Kalluri, Krishna S. Minto, Elise L. Patterson, Jason Zhang, Linying Hripcsak, George Gürsoy, Gamze Elhadad, Noémie Natarajan, Karthik Department of Biomedical Informatics Columbia University Irving Medical Center United States Institute for Informatics Data Science and Biostatistics Washington University St Louis United States Medical Informatics Services New York-Presbyterian Hospital United States Observational Health Data Sciences and Informatics New York Genome Center New YorkNY United States Department of Computer Science Columbia University United States

Synthetic Electronic Health Records (EHR) have emerged as a pivotal tool in advancing healthcare applications and machine learning models, particularly for researchers without direct access to healthcare data. Although existing methods, like rule-based approaches and generative adversarial networks (GANs), generate synthetic data that resembles real-world EHR data, these methods often use a tabular format, disregarding temporal dependencies in patient histories and limiting data replication. Recently, there has been a growing interest in leveraging Generative Pre-trained Transformers (GPT) for EHR data. This enables applications like disease progression analysis, population estimation, counter-factual reasoning, and synthetic data generation. In this work, we focus on synthetic data generation and demonstrate the capability of training a GPT model using a particular patient representation derived from CEHR-BERT, enabling us to generate patient sequences, which can be seamlessly converted to the Observational Medical Outcomes Partnership (OMOP) data format. © 2024, CC BY.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Predicting Protein-Protein Interactions Using Self-Attention-Based Deep Neural Networks and FastText Embeddings

Predicting Protein-Protein Interactions Using Self-Attention...

引用

International Conference on Computing and Networking Technology (ICCNT)

作者： I. R. Oviya N Shanmukha Sravya Kalpana Raja Department of Computer Science Engineering Amrita School Of Computing Amrita Vishwa Vidyapeetham Chennai India Section for Biomedical Informatics and Data Science School of Medicine Yale University New Haven Connecticut USA

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

Predicting Protein-protein interactions are central to understanding the intricate mechanisms governing cellular functions, as proteins seldom act in isolation. While current machine learning and deep learning models focus on encoding protein sequences based on residue composition or physicochemical properties, they often overlook the complex relationships between amino acid segments within lengthy sequences. To address these limitations, we propose a novel hybrid model that combines a self-attention to enhance Deep neural network feature extraction with an with FastText Protein embeddings for predicting PPI. Utilizing transfer learning via the unsupervised FastText embedding technique to derive statistical representations of protein sequences, offers advantages over traditional protein encoding techniques in the context of analysing k-Mer sequences. Our model’s performance is assessed through $\mathbf{5 - f o l d}$ cross-validation on datasets of *** and Human-Bacillus Anthracis protein interactions, resulting in accuracies of $\mathbf{9 2. 2 8 \%}$ and $\mathbf{9 4. 0 8 \%}$, respectively. Compared to existing machine learning models, our model demonstrates superior performance on both datasets.

关键词： Proteins Accuracy Transfer learning Pathogenesis Artificial neural networks Organizations Predictive models Feature extraction Encoding Vectors

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：