Add radiology sentence-level classification example using PyHealth tokenizer and metrics

[GabrielWarner]

Who I am

• Gabriel Warner (gsw3), UIUC DL4H student

Contribution type

• Example notebook for a clinical NLP task

Summary
This PR adds a simple, runnable example demonstrating sentence-level radiology classification (normal / abnormal / uncertain) using PyHealth’s Tokenizer and multiclass_metrics_fn.
The notebook uses a small toy dataset, trains a minimal PyTorch model, and evaluates it end-to-end.
This example relates directly to the sentence classification component of my DL4H final project.

File added

• examples/radiology_sentence_classification_demo.ipynb

How to run

1. Open the notebook
2. Run the setup cell:

   %pip install pyhealth
   
   

Paper Integrating ChatGPT into Secure Hospital Networks: A Case Study on Improving Radiology Report Analysis