In this study, we revisit named entity recognition (NER) in the biomedical domain from a multimodal perspective, with a particular focus on applications in low-resource languages. Existing research primarily relies on unimodal methods for NER, which limits the potential for capturing diverse information. To address this limitation, we propose a novel method that integrates a cross-modal generation module to transform unimodal data into multimodal data, thereby enabling the use of enriched multimodal information for NER. Additionally, we design a cross-modal filtering module to mitigate the adverse effects of text-image mismatches in multimodal NER. We validate our proposed method on two biomedical datasets specifically curated for low-resource languages. Experimental results demonstrate that our method significantly enhances the performance of NER, highlighting its effectiveness and potential for broader applications in biomedical research and low-resource language contexts.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.jbi.2024.104754 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Factors Associated with Nonspecialist Quality of Delivery within a Family Strengthening Intervention in Rwanda: a Parallel Latent Growth Model.
Prev Sci
January 2025
Boston College School of Social Work, 140 Commonwealth Ave, Chestnut Hill, Boston, MA, 02467, USA.
In task-shared, mental health, and psychosocial support interventions, monitoring the quality of delivery (fidelity and competence) of nonspecialist providers is critical. Quality of delivery is frequently reported in brief, summary statistics, and while both fidelity and competence scores tend to be high, rarely have factors associated with quality of delivery in low-resource, mental health, and psychosocial support interventions been examined using inferential statistics. Understanding both modifiable and non-modifiable predictors of quality of delivery is important for adapting training and supervision approaches throughout intervention delivery.
View Article and Find Full Text PDFA comprehensive dataset and neural network approach for named entity recognition in the Uzbek language.
Data Brief
February 2025
Tashkent institute of textile and light industry, 5, Shoxdjaxon str., Tashkent city 100100, Uzbekistan.
In this study, the authors presented a dataset for named entity recognition in the Uzbek language. The dataset consists of 2000 sentences and 25,865 words, and the sources were legal documents and hand-crafted sentences annotated using the BIOES scheme. The study is complemented by the fact that the authors demonstrated the applications of the created dataset by training a language model using the CNN + LSTM architecture, which achieves high accuracy in NER tasks, with an F1 score of 90.
View Article and Find Full Text PDFRed cell distribution width as a cardiovascular risk predictor in adults with hypertension in sub-Saharan Africa.
J Hum Hypertens
January 2025
Department of Pediatrics and Child Health, University of Ilorin, Ilorin, Nigeria.
Article Synopsis
- Red cell distribution width (RDW) is a measure of variability in red blood cell sizes and may indicate cardiovascular disease (CVD) risk, especially in low-resource areas, according to a study in Ghana and Nigeria.
- The study involved 319 adults with hypertension who underwent assessments like blood pressure and RDW measurement, revealing an average RDW of 13.96% and a median CVD risk score of 8.11%.
- Findings showed that RDW is positively correlated with age, systolic and diastolic blood pressure, and WHO CVD risk scores, suggesting RDW could be a useful predictor of CVD risk in this population, particularly with an RDW cutoff of >14
Two-Layer Retrieval-Augmented Generation Framework for Low-Resource Medical Question Answering Using Reddit Data: Proof-of-Concept Study.
J Med Internet Res
January 2025
Department of Biomedical Informatics, School of Medicine, Emory University, Atlanta, GA, United States.
Background: The increasing use of social media to share lived and living experiences of substance use presents a unique opportunity to obtain information on side effects, use patterns, and opinions on novel psychoactive substances. However, due to the large volume of data, obtaining useful insights through natural language processing technologies such as large language models is challenging.
Objective: This paper aims to develop a retrieval-augmented generation (RAG) architecture for medical question answering pertaining to clinicians' queries on emerging issues associated with health-related topics, using user-generated medical information on social media.
Engaging fathers(to-be): a pilot study on the adaptation and programme experience of SMS4baba intervention in Kenya's informal settlements.
BMC Public Health
December 2024
Institute for Human Development, Aga Khan University, Nairobi, Kenya.
Background: Engaging fathers(to-be) can improve maternal, newborn, and child health outcomes. However, father-focused interventions in low-resource settings are under-researched. As part of an integrated early childhood development pilot cluster randomised trial in Nairobi's informal settlements, this study aimed to test the feasibility of a text-only intervention for fathers (SMS4baba) adapted from one developed in Australia (SMS4dads).
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!