BreastVisionAI

• 

  Title and input layout

• 

  Result of the model

• 

  Visualization layout

Inspiration

Breast cancer is one of the leading causes of death among women worldwide, but early detection can significantly improve survival rates. We wanted to leverage AI to build a tool that makes detection faster, accessible, and interactive

What it does

Chatbot that collects 30 diagnostic features step by step. AI-assisted mammogram (X-ray) analysis. Detailed model evaluation with accuracy, recall, precision, and confusion matrix.

How we built it

Dataset preprocessing with Pandas. Machine learning model using LinearSVC. Interactive web app built with Streamlit. X-ray image analysis powered by Gemini AI.

Visualization using Matplotlib and Seaborn

Challenges we ran into

Handling 30 different diagnostic features in a user-friendly way. Designing an intuitive chatbot-like flow for medical data entry. Integrating generative AI for X-ray analysis while keeping results concise and explainable.

Accomplishments that we're proud of

Achieved 97%+ accuracy in cancer detection. Built an interactive, multi-functional Streamlit app. Combined structured data analysis with X-ray image AI in a single platform.

What we learned

How to balance accuracy vs. usability in ML applications. Importance of UI/UX design in healthcare AI tools. Practical integration of ML models with conversational interfaces.

What's next for BreastVisionAI

Expand dataset for higher generalizability. Add deep learning models (CNNs) for more accurate X-ray detection. Deploy as a cloud-based web app for doctors and researchers. Enhance explainability with SHAP/Grad-CAM visualizations.

Built With

• google-gemini-ai
• matplotlib
• numpy
• pandas
• python
• scikit-learn
• seaborn
• streamlit