PCOSaathi : Women's Health Clustering & Personalized Recommendations

📌 Overview

PCOSaathi is an AI-powered health companion designed to support women in managing PCOS (Polycystic Ovary Syndrome) by providing personalized health recommendations. Submitted under the Infosys Springboard iAccelerate Women’s Hackathon 2025.

The application leverages machine learning to analyze women's health data and provide personalized health recommendations.It clusters users based on key health parameters like weight gain, hair loss, acne, period regularity, and exercise, etc to generate insights for managing conditions like PCOS (Polycystic Ovary Syndrome).

The application is built using Streamlit for the frontend, with Agglomerative Clustering and Gaussian Mixture Models (GMM) for clustering. The recommendations are tailored to cluster women based on their health needs and prefrences in context of Indian dietary habits and lifestyle preferences.

🚀 Features

• Data-driven clustering: Uses machine learning models to group users based on health parameters, needs and prefrences.
• Personalized recommendations: Offers customized lifestyle, diet, and exercise advice.
• Indian context: Tailored suggestions based on local food, routines, and wellness practices.
• User-friendly UI: Built with Streamlit for easy interaction.

🏗 Tech Stack

• Frontend: Streamlit
• Backend: Python (Pandas, NumPy, Scikit-Learn)
• Visualization: Seaborn, Matplotlib
• Machine Learning Models:
  • Agglomerative Clustering (Hierarchical Clustering)
  • Gaussian Mixture Model (GMM)
• Data Processing: StandardScaler for normalization

📂 Project Structure

├── Women-Health-Clustering/
│   ├── app.py                    # Main Streamlit application
│   ├── requirements.txt          # Python dependencies
|   ├── myData.py                 #Jupyter Notebook
│   ├── dataset/
│   │   ├── CLEAN-PCOS SURVEY SPREADSHEET.csv    # Cleaned dataset
│   ├── README.md                  # Project documentation

⚡ Installation & Setup

1. Clone the repository

   git clone https://github.com/yourusername/Women-Health-Clustering.git
   cd Women-Health-Clustering

2. Create a virtual environment (Recommended)

   python -m venv venv
   source venv/bin/activate  # On Mac/Linux
   venv\Scripts\activate     # On Windows

3. Install dependencies

   pip install -r requirements.txt

4. Run the application

   streamlit run app.py

🎯 How It Works

1. User inputs their health data: Age, weight gain, hair loss, acne severity, period regularity, and exercise routine.
2. Machine learning models analyze the data: Clusters users based on similarities.
3. Personalized health recommendations: The app provides practical, culturally relevant health advice.

Archetecture Diagram

| |

FlowChart

| |

🖼 Screenshots

| Input Section | Recommendation Output | |-----------|--------------|-----------------------| | | |

🎥 Demo Video

Watch the full demo on YouTube: YouTube Link Here

Future Scope

Here’s a well-defined Future Scope section you can add to your README:

---

🚀 Future Scope

PCOSaathi has the potential to evolve into a comprehensive AI-powered women’s health assistant. Future improvements include:

1️⃣ Enhanced Machine Learning Models

• Integration of deep learning models (LSTMs, Transformer-based models) for more accurate health predictions.
• Personalized recommendations using reinforcement learning to improve with user feedback.

2️⃣ Expansion of Health Parameters

• Addition of hormonal levels, family history, sleep patterns, and mental health factors for deeper analysis.
• Incorporating wearable device data (Fitbit, Apple Health, etc.) to track real-time metrics.

3️⃣ AI-Powered Chatbot for Health Guidance

• Implementing an NLP-based chatbot to provide instant answers to PCOS-related queries.
• Chatbot trained on verified medical resources to ensure accurate responses.

4️⃣ Doctor & Nutritionist Integration

• Feature to connect users with medical experts, gynecologists, and nutritionists based on their health profile.
• Dietitian-approved meal plans tailored to different PCOS types.

5️⃣ Community & Social Support Features

• A peer-support forum where users can share experiences, progress, and tips.
• “Someone Who’s Been There” section where users can ask questions and get advice from women with similar experiences.

6️⃣ Multi-Language & Regional Adaptation

• Supporting Indian regional languages for wider accessibility.
• Dietary recommendations tailored for various Indian cuisines and regional food preferences.

7️⃣ Mobile App Development

• Expanding beyond Streamlit to a Flutter-based mobile app for broader reach.
• Push notifications for reminders, health tips, and AI-generated lifestyle suggestions.

🔥 Contributing

Want to improve this project? Follow these steps:

1. Fork the repository
2. Create a new branch (feature-new-improvement)
3. Make changes and commit (git commit -m 'Add a cool feature')
4. Push to your branch and submit a Pull Request

📜 License

This project is licensed under the MIT License.

---

🌸 Empowering women's health through data-driven insights!