README.md

---
tags:
- lora
- text-generation-inference
- biology
- medical
- ontogpt
widget:
- text: '-'
  output:
    url: images/example.png
base_model: meta-llama/Llama-3.1-8B
instance_prompt: null
license: mit
pipeline_tag: text2text-generation
library_name: peft
---
# IBD MIMIC LLM

<Gallery />

## Model description 


![image.png](https:&#x2F;&#x2F;cdn-uploads.huggingface.co&#x2F;production&#x2F;uploads&#x2F;6631fd0ee97d286a1d314eee&#x2F;6W__2mgq2MeDfoXbSEmeL.png)

This model was developed using 30 manually curated imaging reports from IBD patients in the MIMIC dataset. It is designed to work with OntoGPT to extract and structure key pathological findings from free-text reports. The model assigns weights to specific terms based on their relevance, ensuring more accurate and standardized data extraction. These weights help prioritize important findings and improve the precision of downstream analysis, such as clinical phenotyping or automated report structuring.


## Download model

Weights for this model are available in Safetensors,PyTorch format.
Additional files available: https://github.com/UoS-HGIG/IBD_LLM

[Download](/UoS-HGIG/MIMIC/tree/main) them in the Files & versions tab.

## Reference
 Application of Generative Artificial Intelligence to Utilise Unstructured Clinical Data for Acceleration of Inflammatory Bowel Disease Research. Alex Z Kadhim, Zachary Green, Iman Nazari, Jonathan Baker, Michael George, Ashley Heinson, Matt Stammers, Christopher Kipps, R Mark Beattie, James J Ashton, Sarah Ennis. medRxiv 2025.03.07.25323569; doi: https://doi.org/10.1101/2025.03.07.25323569