E3-JSI/gliner-multi-med-ner-synthetic-v1

synthetic-multi-med-notes-ner-gliner_multi-v2.1

A fine-tuned GLiNER v2.1 model for Named Entity Recognition (NER) in medical notes, trained on multilingual synthetic data.

Model Details

• Task: Named Entity Recognition (NER)
• Languages: English, Italian, Spanish, German, French, Dutch, Greek, Portuguese, Slovenian

Trained Label Set

• Comorbidity
• Condition
• Date
• Device
• Drug
• Drug dose
• Event
• Measurement
• Observation
• Operation
• Procedure
• Rehabilitation
• Specimen
• Symptom
• Test
• Test score
• Treatment
• Treatment complication
• Visit

Usage

from gliner import GLiNER

model = GLiNER.from_pretrained("E3-JSI/gliner-multi-med-ner-synthetic-v1")

text = (
    "On 2026-03-15, the patient visited cardiology with chest pain and fatigue."
    "ECG and troponin test were ordered. BP 150/95 mmHg, HbA1c 8.2%."
    "Diagnosed with hypertension and type 2 diabetes with obesity as comorbidity."
    "Started metformin 500 mg twice daily and amlodipine 5 mg daily."
    "Planned cardiac catheterization procedure and referral to rehabilitation."
)

labels = [
    "Comorbidity", "Condition", "Date", "Device", "Drug", "Drug dose", "Event",
    "Measurement", "Observation", "Operation", "Procedure", "Rehabilitation",
    "Specimen", "Symptom", "Test", "Test score", "Treatment",
    "Treatment complication", "Visit"
]

entities = model.predict_entities(text, labels, threshold=0.5)
for entity in entities:
    print(f"{entity['text']} -> {entity['label']} ({entity['score']:.3f})")

Training Details

• Synthetic Data: Multilingual medical notes generated from templates
• Training Quality: High-quality synthetic annotations for improved generalization

License

Licensed under Apache 2.0

Aknowledgements

Funded by the European Union. UK participants in Horizon Europe Project PREPARE are supported by UKRI grant number 10086219 (Trilateral Research). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Health and Digital Executive Agency (HADEA) or UKRI. Neither the European Union nor the granting authority nor UKRI can be held responsible for them. Grant Agreement 101080288 PREPARE HORIZON-HLTH-2022-TOOL-12-01.