B

Bsc Bio Ehr Es

Developed by PlanTL-GOB-ES
Pre-trained language model optimized for Spanish biomedical and clinical texts, supporting clinical NLP tasks
Downloads 624
Release Time : 4/8/2022

Model Overview

Spanish biomedical domain-specific model based on RoBERTa architecture, trained on mixed biomedical corpus and real clinical records, suitable for clinical text analysis tasks

Model Features

Domain-specific optimization
Trained on 1 billion tokens of biomedical-clinical mixed corpus, including 278,000 real clinical documents
Multi-source data integration
Integrates 11 professional data sources including medical crawlers, clinical cases, and electronic health records
Clinical NER advantage
Outperforms general and multilingual models on clinical NER tasks such as PharmaCoNER and CANTEMIST

Model Capabilities

Biomedical text understanding
Clinical entity recognition
Electronic health record analysis
Medical text classification

Use Cases

Clinical information extraction
Drug name recognition
Identify chemical drug mentions from clinical texts
Achieved 0.8913 F1 score on PharmaCoNER task
Tumor morphology recognition
Identify Spanish oncological terminology
Achieved 0.8340 F1 score on CANTEMIST task
Electronic health record processing
Discharge report analysis
Parse clinical variables in stroke patient discharge reports
Achieved 0.8756 F1 score on ICTUSnet dataset
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase