YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Model Card: Medical Referral Information Extraction and Severity Prioritization LLM Model Name
Model Description
This model is a fine-tuned Large Language Model (LLM) designed to extract critical patient information from medical referral documents and categorize or prioritize patients based on the severity of their condition. The model identifies structured entities such as patient demographics, medical history, symptoms, diagnoses, and referral details. It can also provide a severity ranking or urgency score to help healthcare providers prioritize cases efficiently.
Key Features
Extracts patient demographic details (name, age, gender, contact, etc.)
Identifies clinical information such as symptoms, past medical history, medications, allergies
Extracts diagnosis or suspected conditions from referral notes
Assigns severity scores or categories (high, medium, low) based on extracted information
Outputs structured data suitable for integration into electronic health records (EHRs) or triage systems
Intended Use
Primary Use: Automating information extraction from medical referrals to reduce manual effort and improve triage efficiency.
Secondary Use: Assisting healthcare providers in prioritizing patient cases based on severity.
Users: Hospitals, clinics, medical administrative staff, health tech developers.
Limitations
Performance may vary depending on the format and quality of referral documents.
Not a substitute for clinical judgment; severity scoring should be validated by a healthcare professional.
May not generalize well to languages, medical terminologies, or healthcare systems not included in the training data.
Privacy considerations must be observed when using real patient data.
Training Data
Fine-tuned on anonymized medical referral texts with labeled entities and severity tags.
Data included diverse clinical scenarios to improve robustness, but the model is optimized primarily for English-language referrals.
No personally identifiable information (PII) was used.
- Downloads last month
- 6