clinicalnerpt-disease
| Property | Value |
|---|---|
| Author | PUCPR |
| Framework | PyTorch |
| Training Data | SemClinBr |
| Language | Portuguese |
What is clinicalnerpt-disease?
clinicalnerpt-disease is a specialized Named Entity Recognition (NER) model designed to identify disease mentions in Portuguese clinical texts. It's part of the comprehensive BioBERTpt project, which includes 13 different clinical entity recognition models compatible with UMLS (Unified Medical Language System). The model was trained on the Brazilian clinical corpus SemClinBr using the BioBERTpt architecture with 10 epochs in IOB2 format.
Implementation Details
The model leverages transformer architecture and was developed using PyTorch. It builds upon the BioBERTpt base model, which was created by transfer learning from multilingual-BERT and fine-tuned on Portuguese clinical and biomedical texts.
- Trained specifically for disease entity recognition in clinical texts
- Uses IOB2 tagging format for entity identification
- Optimized through 10 training epochs
- Built on BioBERTpt architecture
Core Capabilities
- Accurate identification of disease mentions in Portuguese clinical texts
- Compatible with UMLS semantic types
- Supports real-time inference
- Handles complex clinical narratives and medical terminology
Frequently Asked Questions
Q: What makes this model unique?
This model is specifically optimized for Portuguese clinical text analysis, addressing the gap in non-English medical NLP capabilities. It's part of a larger ecosystem of clinical NER models and has been trained on authentic Brazilian medical texts.
Q: What are the recommended use cases?
The model is ideal for automated processing of Portuguese clinical documents, electronic health records analysis, medical research, and clinical decision support systems where disease entity recognition is crucial.
