Update README.md

README.md

@@ -1,13 +1,49 @@
+# 🧠 Stroke Pain Detection App
+
+A Streamlit-based app for automatically assessing pain intensity in stroke patients using facial asymmetry and deep learning.
+
+This project uses three steps:
+1. **Detect the affected side** of a stroke patient's face using a Keras-based CNN model.
+2. **Crop the unaffected half** of the face.
+3. Use a PyTorch model to **predict the pain intensity (PSPI score)** from the expressive half.
+
 ---
-title: StrokePatientsModel
-emoji: 🏢
-colorFrom: gray
-colorTo: gray
-sdk: streamlit
-sdk_version: 1.44.1
-app_file: app.py
-pinned: false
-license: mit
+
+## 🚀 How to Use This App
+
+Simply upload a **full-face image** of a stroke patient and let the model:
+- Detect the affected side
+- Crop the unaffected side
+- Predict the patient's pain intensity (PSPI)
+
+✅ You can access the live app here:  
+[🔗 Launch App →](https://huggingface.co/spaces/AdhamQQ/StrokePatientsModel)
+
+
+## 💾 Models Used
+
+These models are automatically downloaded on first run:
+
+| Model                             | Description                            | Link                                                                 |
+|----------------------------------|----------------------------------------|----------------------------------------------------------------------|
+| `cnn_stroke_model.keras`         | Detects the affected side of the face  | [Download](https://huggingface.co/AdhamQQ/cnn_stroke_model/resolve/main/cnn_stroke_model.keras)         |
+| `right_side_pain_model.pth`      | Predicts pain intensity (PSPI score)   | [Download](https://huggingface.co/AdhamQQ/cnn_stroke_model/resolve/main/right_side_pain_model.pth)      |
+
+---
+
+## 🧠 Project Goals
+
+This app is part of a larger system to:
+- Automatically detect which side of a stroke patient's face is impaired
+- Use the **expressive side** to analyze pain using PSPI (Prkachin and Solomon Pain Intensity) scores
+
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+## ✨ Credits
+
+using:
+- Streamlit
+- TensorFlow/Keras
+- PyTorch
+- Hugging Face Hub
+