Biomedical Question Answering Research Articles

Optimizing biomedical information retrieval with a keyword frequency-driven prompt enhancement strategy

BackgroundMining the vast pool of biomedical literature to extract accurate responses and relevant references is challenging due to the domain's interdisciplinary nature, specialized jargon, and continuous evolution. Early natural language processing (NLP) approaches often led to incorrect answers as they failed to comprehend the nuances of natural language. However, transformer models have significantly advanced the field by enabling the creation of large language models (LLMs), enhancing question-answering (QA) tasks. Despite these advances, current LLM-based solutions for specialized domains like biology and biomedicine still struggle to generate up-to-date responses while avoiding “hallucination” or generating plausible but factually incorrect responses.ResultsOur work focuses on enhancing prompts using a retrieval-augmented architecture to guide LLMs in generating meaningful responses for biomedical QA tasks. We evaluated two approaches: one relying on text embedding and vector similarity in a high-dimensional space, and our proposed method, which uses explicit signals in user queries to extract meaningful contexts. For robust evaluation, we tested these methods on 50 specific and challenging questions from diverse biomedical topics, comparing their performance against a baseline model, BM25. Retrieval performance of our method was significantly better than others, achieving a median Precision@10 of 0.95, which indicates the fraction of the top 10 retrieved chunks that are relevant. We used GPT-4, OpenAI's most advanced LLM to maximize the answer quality and manually accessed LLM-generated responses. Our method achieved a median answer quality score of 2.5, surpassing both the baseline model and the text embedding-based approach. We developed a QA bot, WeiseEule (https://github.com/wasimaftab/WeiseEule-LocalHost), which utilizes these methods for comparative analysis and also offers advanced features for review writing and identifying relevant articles for citation.ConclusionsOur findings highlight the importance of prompt enhancement methods that utilize explicit signals in user queries over traditional text embedding-based approaches to improve LLM-generated responses for specialized queries in specialized domains such as biology and biomedicine. By providing users complete control over the information fed into the LLM, our approach addresses some of the major drawbacks of existing web-based chatbots and LLM-based QA systems, including hallucinations and the generation of irrelevant or outdated responses.

Enhancing Biomedical Question Answering with Large Language Models

In the field of Information Retrieval, biomedical question answering is a specialized task that focuses on answering questions related to medical and healthcare domains. The goal is to provide accurate and relevant answers to the posed queries related to medical conditions, treatments, procedures, medications, and other healthcare-related topics. Well-designed models should efficiently retrieve relevant passages. Early retrieval models can quickly retrieve passages but often with low precision. In contrast, recently developed Large Language Models can retrieve documents with high precision but at a slower pace. To tackle this issue, we propose a two-stage retrieval approach that initially utilizes BM25 for a preliminary search to identify potential candidate documents; subsequently, a Large Language Model is fine-tuned to evaluate the relevance of query–document pairs. Experimental results indicate that our approach achieves comparative performances on the BioASQ and the TREC-COVID datasets.

Efficient Machine Reading Comprehension for Health Care Applications: Algorithm Development and Validation of a Context Extraction Approach.

Extractive methods for machine reading comprehension (MRC) tasks have achieved comparable or better accuracy than human performance on benchmark data sets. However, such models are not as successful when adapted to complex domains such as health care. One of the main reasons is that the context that the MRC model needs to process when operating in a complex domain can be much larger compared with an average open-domain context. This causes the MRC model to make less accurate and slower predictions. A potential solution to this problem is to reduce the input context of the MRC model by extracting only the necessary parts from the original context. This study aims to develop a method for extracting useful contexts from long articles as an additional component to the question answering task, enabling the MRC model to work more efficiently and accurately. Existing approaches to context extraction in MRC are based on sentence selection strategies, in which the models are trained to find the sentences containing the answer. We found that using only the sentences containing the answer was insufficient for the MRC model to predict correctly. We conducted a series of empirical studies and observed a strong relationship between the usefulness of the context and the confidence score output of the MRC model. Our investigation showed that a precise input context can boost the prediction correctness of the MRC and greatly reduce inference time. We proposed a method to estimate the utility of each sentence in a context in answering the question and then extract a new, shorter context according to these estimations. We generated a data set to train 2 models for estimating sentence utility, based on which we selected more precise contexts that improved the MRC model's performance. We demonstrated our approach on the Question Answering Data Set for COVID-19 and Biomedical Semantic Indexing and Question Answering data sets and showed that the approach benefits the downstream MRC model. First, the method substantially reduced the inference time of the entire question answering system by 6 to 7 times. Second, our approach helped the MRC model predict the answer more correctly compared with using the original context (F1-score increased from 0.724 to 0.744 for the Question Answering Data Set for COVID-19 and from 0.651 to 0.704 for the Biomedical Semantic Indexing and Question Answering). We also found a potential problem where extractive transformer MRC models predict poorly despite being given a more precise context in some cases. The proposed context extraction method allows the MRC model to achieve improved prediction correctness and a significantly reduced MRC inference time. This approach works technically with any MRC model and has potential in tasks involving processing long texts.

Document Retrieval System for Biomedical Question Answering

In this paper, we describe our biomedical document retrieval system and answers extraction module, which is part of the biomedical question answering system. Approximately 26.5 million PubMed articles are indexed as a corpus with the Apache Lucene text search engine. Our proposed system consists of three parts. The first part is the question analysis module, which analyzes the question and enriches it with biomedical concepts related to its wording. The second part of the system is the document retrieval module. In this step, the proposed system is tested using different information retrieval models, like the Vector Space Model, Okapi BM25, and Query Likelihood. The third part is the document re-ranking module, which is responsible for re-arranging the documents retrieved in the previous step. For this study, we tested our proposed system with 6B training questions from the BioASQ challenge task. We obtained the best MAP score on the document retrieval phase when we used Query Likelihood with the Dirichlet Smoothing model. We used the sequential dependence model at the re-rank phase, but this model produced a worse MAP score than the previous phase. In similarity calculation, we included the Named Entity Recognition (NER), UMLS Concept Unique Identifiers (CUI), and UMLS Semantic Types of the words in the question to find the sentences containing the answer. Using this approach, we observed a performance enhancement of roughly 25% for the top 20 outcomes, surpassing another method employed in this study, which relies solely on textual similarity.

Taiyi: a bilingual fine-tuned large language model for diverse biomedical tasks.

Most existing fine-tuned biomedical large language models (LLMs) focus on enhancing performance in monolingual biomedical question answering and conversation tasks. To investigate the effectiveness of the fine-tuned LLMs on diverse biomedical natural language processing (NLP) tasks in different languages, we present Taiyi, a bilingual fine-tuned LLM for diverse biomedical NLP tasks. We first curated a comprehensive collection of 140 existing biomedical text mining datasets (102 English and 38 Chinese datasets) across over 10 task types. Subsequently, these corpora were converted to the instruction data used to fine-tune the general LLM. During the supervised fine-tuning phase, a 2-stage strategy is proposed to optimize the model performance across various tasks. Experimental results on 13 test sets, which include named entity recognition, relation extraction, text classification, and question answering tasks, demonstrate that Taiyi achieves superior performance compared to general LLMs. The case study involving additional biomedical NLP tasks further shows Taiyi's considerable potential for bilingual biomedical multitasking. Leveraging rich high-quality biomedical corpora and developing effective fine-tuning strategies can significantly improve the performance of LLMs within the biomedical domain. Taiyi shows the bilingual multitasking capability through supervised fine-tuning. However, those tasks such as information extraction that are not generation tasks in nature remain challenging for LLM-based generative approaches, and they still underperform the conventional discriminative approaches using smaller language models.

Evaluating the ChatGPT family of models for biomedical reasoning and classification.

Large language models (LLMs) have shown impressive ability in biomedical question-answering, but have not been adequately investigated for more specific biomedical applications. This study investigates ChatGPT family of models (GPT-3.5, GPT-4) in biomedical tasks beyond question-answering. We evaluated model performance with 11 122 samples for two fundamental tasks in the biomedical domain-classification (n = 8676) and reasoning (n = 2446). The first task involves classifying health advice in scientific literature, while the second task is detecting causal relations in biomedical literature. We used 20% of the dataset for prompt development, including zero- and few-shot settings with and without chain-of-thought (CoT). We then evaluated the best prompts from each setting on the remaining dataset, comparing them to models using simple features (BoW with logistic regression) and fine-tuned BioBERT models. Fine-tuning BioBERT produced the best classification (F1: 0.800-0.902) and reasoning (F1: 0.851) results. Among LLM approaches, few-shot CoT achieved the best classification (F1: 0.671-0.770) and reasoning (F1: 0.682) results, comparable to the BoW model (F1: 0.602-0.753 and 0.675 for classification and reasoning, respectively). It took 78 h to obtain the best LLM results, compared to 0.078 and 0.008 h for the top-performing BioBERT and BoW models, respectively. The simple BoW model performed similarly to the most complex LLM prompting. Prompt engineering required significant investment. Despite the excitement around viral ChatGPT, fine-tuning for two fundamental biomedical natural language processing tasks remained the best strategy.

Benchmarking Large Language Models in Evidence-Based Medicine.

Evidence-based medicine (EBM) represents a paradigm of providing patient care grounded in the most current and rigorously evaluated research. Recent advances in large language models (LLMs) offer a potential solution to transform EBM by automating labor-intensive tasks and thereby improving the efficiency of clinical decision-making. This study explores integrating LLMs into the key stages in EBM, evaluating their ability across evidence retrieval (PICO extraction, biomedical question answering), synthesis (summarizing randomized controlled trials), and dissemination (medical text simplification). We conducted a comparative analysis of seven LLMs, including both proprietary and open-source models, as well as those fine-tuned on medical corpora. Specifically, we benchmarked the performance of various LLMs on each EBM task under zero-shot settings as baselines, and employed prompting techniques, including in-context learning, chain-of-thought reasoning, and knowledge-guided prompting to enhance their capabilities. Our extensive experiments revealed the strengths of LLMs, such as remarkable understanding capabilities even in zero-shot settings, strong summarization skills, and effective knowledge transfer via prompting. Promoting strategies such as knowledge-guided prompting proved highly effective (e.g., improving the performance of GPT-4 by 13.10% over zero-shot in PICO extraction). However, the experiments also showed limitations, with LLM performance falling well below state-of-the-art baselines like PubMedBERT in handling named entity recognition tasks. Moreover, human evaluation revealed persisting challenges with factual inconsistencies and domain inaccuracies, underscoring the need for rigorous quality control before clinical application. This study provides insights into enhancing EBM using LLMs while highlighting critical areas for further research. The code is publicly available on Github.

Developing ChatGPT for biology and medicine: a complete review of biomedical question answering.

ChatGPT explores a strategic blueprint of question answering (QA) to deliver medical diagnoses, treatment recommendations, and other healthcare support. This is achieved through the increasing incorporation of medical domain data via natural language processing (NLP) and multimodal paradigms. By transitioning the distribution of text, images, videos, and other modalities from the general domain to the medical domain, these techniques have accelerated the progress of medical domain question answering (MDQA). They bridge the gap between human natural language and sophisticated medical domain knowledge or expert-provided manual annotations, handling large-scale, diverse, unbalanced, or even unlabeled data analysis scenarios in medical contexts. Central to our focus is the utilization of language models and multimodal paradigms for medical question answering, aiming to guide the research community in selecting appropriate mechanisms for their specific medical research requirements. Specialized tasks such as unimodal-related question answering, reading comprehension, reasoning, diagnosis, relation extraction, probability modeling, and others, as well as multimodal-related tasks like vision question answering, image captioning, cross-modal retrieval, report summarization, and generation, are discussed in detail. Each section delves into the intricate specifics of the respective method under consideration. This paper highlights the structures and advancements of medical domain explorations against general domain methods, emphasizing their applications across different tasks and datasets. It also outlines current challenges and opportunities for future medical domain research, paving the way for continued innovation and application in this rapidly evolving field. This comprehensive review serves not only as an academic resource but also delineates the course for future probes and utilization in the field of medical question answering.

A self-supervised language model selection strategy for biomedical question answering

Large neural-based Pre-trained Language Models (PLM) have recently gained much attention due to their noteworthy performance in many downstream Information Retrieval (IR) and Natural Language Processing (NLP) tasks. PLMs can be categorized as either general-purpose, which are trained on resources such as large-scale Web corpora, and domain-specific which are trained on in-domain or mixed-domain corpora. While domain-specific PLMs have shown promising performance on domain-specific tasks, they are significantly more computationally expensive compared to general-purpose PLMs as they have to be either retrained or trained from scratch. The objective of our work in this paper is to explore whether it would be possible to leverage general-purpose PLMs to show competitive performance to domain-specific PLMs without the need for expensive retraining of the PLMs for domain-specific tasks. By focusing specifically on the recent BioASQ Biomedical Question Answering task, we show how different general-purpose PLMs show synergistic behaviour in terms of performance, which can lead to overall notable performance improvement when used in tandem with each other. More concretely, given a set of general-purpose PLMs, we propose a self-supervised method for training a classifier that systematically selects the PLM that is most likely to answer the question correctly on a per-input basis. We show that through such a selection strategy, the performance of general-purpose PLMs can become competitive with domain-specific PLMs while remaining computationally light since there is no need to retrain the large language model itself. We run experiments on the BioASQ dataset, which is a large-scale biomedical question-answering benchmark. We show that utilizing our proposed selection strategy can show statistically significant performance improvements on general-purpose language models with an average of 16.7% when using only lighter models such as DistilBERT and DistilRoBERTa, as well as 14.2% improvement when using relatively larger models such as BERT and RoBERTa and so, their performance become competitive with domain-specific large language models such as PubMedBERT.

Enhancing Biomedical ReQA With Adversarial Hard In-Batch Negative Samples.

Question answering (QA) plays a vital role in biomedical natural language processing. Among question answering tasks, the retrieval question answering (ReQA) aims to directly retrieve the correct answer from candidates and has attracted much attention in the community for its efficiency. Recently, researchers have introduced ReQA into the biomedical domain as BioReQA. Typically BioReQA models rely on the dual-encoder to gain semantic representation and are trained following the settings of dense retrieval. However, they normally utilize easy in-batch negative samples in training process to avoid the extra forwarding cost and GPU memory required by encoding additional negative samples. However, hard negative samples have been proved more important with regard to the overall performance of BioReQA tasks. Therefore in this research, we focus on effectively constructing hard in-batch negative samples. Inspired by the classic linear assignment problem, we propose an Iterative Linear Assignment Grouping (ILAG) algorithm to construct hard in-batch negative samples. To further enhance performance for given hard batches in a low-resource scenario, we also employ adversarial training to augment the difficulty of batches. Extensive experiments have shown our proposed method's promising potential in the area of biomedical retrieval question answering.

A novel self-attention enriching mechanism for biomedical question answering

A novel self-attention enriching mechanism for biomedical question answering

Selective UMLS knowledge infusion for biomedical question answering

One of the artificial intelligence applications in the biomedical field is knowledge-intensive question-answering. As domain expertise is particularly crucial in this field, we propose a method for efficiently infusing biomedical knowledge into pretrained language models, ultimately targeting biomedical question-answering. Transferring all semantics of a large knowledge graph into the entire model requires too many parameters, increasing computational cost and time. We investigate an efficient approach that leverages adapters to inject Unified Medical Language System knowledge into pretrained language models, and we question the need to use all semantics in the knowledge graph. This study focuses on strategies of partitioning knowledge graph and either discarding or merging some for more efficient pretraining. According to the results of three biomedical question answering finetuning datasets, the adapters pretrained on semantically partitioned group showed more efficient performance in terms of evaluation metrics, required parameters, and time. The results also show that discarding groups with fewer concepts is a better direction for small datasets, and merging these groups is better for large dataset. Furthermore, the metric results show a slight improvement, demonstrating that the adapter methodology is rather insensitive to the group formulation.

BioASQ-QA: A manually curated corpus for Biomedical Question Answering

The BioASQ question answering (QA) benchmark dataset contains questions in English, along with golden standard (reference) answers and related material. The dataset has been designed to reflect real information needs of biomedical experts and is therefore more realistic and challenging than most existing datasets. Furthermore, unlike most previous QA benchmarks that contain only exact answers, the BioASQ-QA dataset also includes ideal answers (in effect summaries), which are particularly useful for research on multi-document summarization. The dataset combines structured and unstructured data. The materials linked with each question comprise documents and snippets, which are useful for Information Retrieval and Passage Retrieval experiments, as well as concepts that are useful in concept-to-text Natural Language Generation. Researchers working on paraphrasing and textual entailment can also measure the degree to which their methods improve the performance of biomedical QA systems. Last but not least, the dataset is continuously extended, as the BioASQ challenge is running and new data are generated.

Improving Biomedical Question Answering by Data Augmentation and Model Weighting.

Biomedical Question Answering aims to extract an answer to the given question from a biomedical context. Due to the strong professionalism of specific domain, it's more difficult to build large-scale datasets for specific domain question answering. Existing methods are limited by the lack of training data, and the performance is not as good as in open-domain settings, especially degrading when facing to the adversarial sample. We try to resolve the above issues. First, effective data augmentation strategies are adopted to improve the model training, including slide window, summarization and round-trip translation. Second, we propose a model weighting strategy for the final answer prediction in biomedical domain, which combines the advantage of two models, open-domain model QANet and BioBERT pre-trained in biomedical domain data. Finally, we give adversarial training to reinforce the robustness of the model. The public biomedical dataset collected from PubMed provided by BioASQ challenge is used to evaluate our approach. The results show that the model performance has been improved significantly compared to the single model and other models participated in BioASQ challenge. It can learn richer semantic expression from data augmentation and adversarial samples, which is beneficial to solve more complex question answering problems in biomedical domain.

Survey on the Biomedical Text Summarization Techniques with an Emphasis on Databases, Techniques, Semantic Approaches, Classification Techniques, and Similarity Measures

Biomedical text summarization (BTS) is proving to be an emerging area of work and research with the need for sustainable healthcare applications such as evidence-based medicine practice (EBM) and telemedicine which help effectively support healthcare needs of the society. However, with the rapid growth in the biomedical literature and the diversities in its structure and resources, it is becoming challenging to carry out effective text summarization for better insights. The goal of this work is to conduct a comprehensive systematic literature review of significant and high-impact literary work in BTS with a deep understanding of its major artifacts such as databases, semantic similarity measures, and semantic enrichment approaches. In the systematic literature review conducted, we applied search filters to find high-impact literature in the biomedical text summarization domain from IEEE, SCOPUS, Elsevier, EBSCO, and PubMed databases. The systematic literature review (SLR) yielded 81 works; those were analyzed for qualitative study. The in-depth study of the literature shows the relevance and efficacy of the deep learning (DL) approach, context-aware feature extraction techniques, and their relevance in BTS. Biomedical question answering (BQA) system is one of the most popular applications of text summarizations for building self-sufficient healthcare systems and are pointing to future research directions. The review culminates in realization of a proposed framework for the BQA system MEDIQA with design of better heuristics for content screening, document screening, and relevance ranking. The presented framework provides an evidence-based biomedical question answering model and text summarizer that can lead to real-time evidence-based clinical support system to healthcare practitioners.

Adversarial Knowledge Distillation Based Biomedical Factoid Question Answering.

Biomedical factoid question answering is an essential application for biomedical information sharing. Recently, neural network based approaches have shown remarkable performance for this task. However, due to the scarcity of annotated data which requires intensive knowledge of expertise, training a robust model on limited-scale biomedical datasets remains a challenge. Previous works solve this problem by introducing useful knowledge. It is found that the interaction between question and answer (QA-interaction) is also a kind of knowledge which could help extract answer accurately. This research develops a knowledge distillation framework for biomedical factoid question answering, in which a teacher model as the knowledge source of QA-interaction is designed to enhance the student model. In addition, to further alleviate the problem of limited-scale dataset, a novel adversarial knowledge distillation technique is proposed to robustly distill the knowledge from teacher model to student model by constructing perturbed examples as additional training data. By forcing the student model to mimic the predicted distributions of teacher model on both original examples and perturbed examples, the knowledge of QA-interaction can be learned by student model. We evaluate the proposed framework on the widely used BioASQ datasets, and experimental results have shown the proposed method's promising potential.

Sequence tagging for biomedical extractive question answering.

MotivationCurrent studies in extractive question answering (EQA) have modeled the single-span extraction setting, where a single answer span is a label to predict for a given question-passage pair. This setting is natural for general domain EQA as the majority of the questions in the general domain can be answered with a single span. Following general domain EQA models, current biomedical EQA (BioEQA) models utilize the single-span extraction setting with post-processing steps.ResultsIn this article, we investigate the question distribution across the general and biomedical domains and discover biomedical questions are more likely to require list-type answers (multiple answers) than factoid-type answers (single answer). This necessitates the models capable of producing multiple answers for a question. Based on this preliminary study, we propose a sequence tagging approach for BioEQA, which is a multi-span extraction setting. Our approach directly tackles questions with a variable number of phrases as their answer and can learn to decide the number of answers for a question from training data. Our experimental results on the BioASQ 7b and 8b list-type questions outperformed the best-performing existing models without requiring post-processing steps.Availability and implementationSource codes and resources are freely available for download at https://github.com/dmis-lab/SeqTagQA.Supplementary information Supplementary data are available at Bioinformatics online.

Ensemble-based Methods for Multi-label Classification on Biomedical Question-Answer Data

Background: Question-answer (QA) is a popular method to seek health-related information and biomedical data. Such questions can refer to more than one medical entity (multi-label) so determining the correct tags is not easy. The question classification (QC) mechanism in a QA system can narrow down the answers we are seeking. Objective: This study develops a multi-label classification using the heterogeneous ensembles method to improve accuracy in biomedical data with long text dimensions. Methods: We used the ensemble method with heterogeneous deep learning and machine learning for multi-label extended text classification. There are 15 various single models consisting of three deep learning (CNN, LSTM, and BERT) and four machine learning algorithms (SVM, kNN, Decision Tree, and Naïve Bayes) with various text representations (TF-IDF, Word2Vec, and FastText). We used the bagging approach with a hard voting mechanism for the decision-making. Results: The result shows that deep learning is more powerful than machine learning as a single multi-label biomedical data classification method. Moreover, we found that top-three was the best number of base learners by combining the ensembles method. Heterogeneous-based ensembles with three learners resulted in an F1-score of 82.3%, which is better than the best single model by CNN with an F1-score of 80%. Conclusion: A multi-label classification of biomedical QA using ensemble models is better than single models. The result shows that heterogeneous ensembles are more potent than homogeneous ensembles on biomedical QA data with long text dimensions. Keywords: Biomedical Question Classification, Ensemble Method, Heterogeneous Ensembles, Multi-Label Classification, Question Answering

An Efficient Method for Biomedical Entity Linking Based on Inter- and Intra-Entity Attention

Biomedical entity linking is an important research problem for many downstream tasks, such as biomedical intelligent question answering, information retrieval, and information extraction. Biomedical entity linking is the task of mapping mentions in medical texts to standard entities in a given knowledge base. Recently, BERT-based models have achieved state-of-the-art results on the biomedical entity linking task. Although this type of method is effective, it brings challenges for fine-tuning and online services in practical industries due to a large number of model parameters and long inference time. In addition, due to the numerous surface variants of biomedical mentions, it is difficult for a single matching module to achieve good results. To address the challenge, we propose an efficient biomedical entity linking method that integrates inter- and intra-entity attention to better capture the information between medical entity mentions and candidate entities themselves and each other, and the model in this paper is more lightweight. Experimental results show that our method achieves competitive performance on two biomedical benchmark datasets, NCBI and ADR, with an accuracy rate of 91.28% and 93.13%, respectively. Moreover, it also achieves comparable or even better results compared to the BERT-based entity linking method while having far fewer model parameters and very high inference speed.

SentiMedQAer: A Transfer Learning-Based Sentiment-Aware Model for Biomedical Question Answering.

Recent advances have witnessed a trending application of transfer learning in a broad spectrum of natural language processing (NLP) tasks, including question answering (QA). Transfer learning allows a model to inherit domain knowledge obtained from an existing model that has been sufficiently pre-trained. In the biomedical field, most QA datasets are limited by insufficient training examples and the presence of factoid questions. This study proposes a transfer learning-based sentiment-aware model, named SentiMedQAer, for biomedical QA. The proposed method consists of a learning pipeline that utilizes BioBERT to encode text tokens with contextual and domain-specific embeddings, fine-tunes Text-to-Text Transfer Transformer (T5), and RoBERTa models to integrate sentiment information into the model, and trains an XGBoost classifier to output a confidence score to determine the final answer to the question. We validate SentiMedQAer on PubMedQA, a biomedical QA dataset with reasoning-required yes/no questions. Results show that our method outperforms the SOTA by 15.83% and a single human annotator by 5.91%.