config.json

---
title: Aphasia Classifier
emoji: 🧠
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: "4.44.0"
app_file: app.py
pinned: false
license: mit
short_description: AI-powered aphasia severity classification using fine-tuned BioBERT
tags:
- medical
- nlp
- aphasia
- biobert
- classification
- speech-therapy
- healthcare
models:
- dmis-lab/biobert-base-cased-v1.1
---

# Aphasia Classifier 🧠

An AI-powered application for classifying aphasia severity levels using a fine-tuned BioBERT model.

## Features

- **Speech Analysis Pipeline**: Text → CHA Format → JSON → BioBERT Classification
- **Severity Classification**: Normal, Mild, Moderate, Severe aphasia levels
- **Confidence Scoring**: Detailed probability distributions for each class
- **Professional Interface**: Medical-grade UI with multiple output views
- **Real-time Processing**: Complete analysis in seconds

## How to Use

1. Enter a speech sample in the text area
2. Click "Analyze Speech" to process through the pipeline
3. View results across multiple tabs:
   - **Results**: Formatted analysis with confidence scores
   - **CHA Format**: Clinical CHAT format output
   - **JSON Data**: Structured data representation
   - **Raw Classification**: Complete model output

## Model Information

- **Base Model**: BioBERT (dmis-lab/biobert-base-cased-v1.1)
- **Fine-tuning**: Specialized for aphasia severity classification
- **Input**: Natural language speech samples
- **Output**: 4-class severity classification with confidence scores

## Disclaimer

⚠️ This tool is for research and educational purposes only. It should not be used as a substitute for professional medical diagnosis or treatment. Always consult with qualified healthcare professionals for medical advice.

## Technical Pipeline

```
Text Input → CHA Formatting → JSON Structure → BioBERT Model → Classification Results
```

Built with Gradio and Transformers library.