Scancer is an AI-driven diagnostic tool designed to assist healthcare professionals in detecting breast cancer using ultrasound images and cell nuclei measurements. By integrating deep learning and traditional machine learning techniques, Scancer delivers fast, accurate classifications of breast tumors as benign or malignant through an intuitive, web-based interface.
-
Dual Input Detection
Combines image-based (ultrasound) and numeric (tumor measurements) data. -
AI Models in Action
Uses CNN, ResNet, Logistic Regression, and Decision Trees for robust diagnosis. -
User-Friendly Interface
Simple, guided workflow designed for use by clinicians and researchers. -
Graphical Insights
Dynamic tumor surface area graphs aid understanding of input data. -
Instant Results
Real-time diagnostic feedback based on combined data inputs. -
Accessibility First
Built with WCAG standards to support all users.
Scancer integrates the strengths of multiple models to ensure accuracy:
- CNN (Convolutional Neural Network) – Analyzes ultrasound images to detect patterns indicative of cancer.
- ResNet – Deep residual learning for enhanced image classification performance.
- Logistic Regression – Provides statistical analysis on clinical measurements.
- Decision Tree – Offers a transparent and interpretable classification process.
These models work in tandem to analyze data, increasing diagnostic confidence through ensemble learning.
Scancer was developed with feedback from healthcare practitioners, emphasizing:
- Clear input and output formats
- Visual interpretation aids
- Accessible design for diverse users
- Real-world validation via user acceptance testing (UAT)
- SSL Encryption – All data transmissions are secured.
- Data Validation – Strict input checks to prevent errors.
- Privacy First – Designed with ethical handling of medical data in mind.
- Improving model accuracy with larger, diverse datasets.
- Enhancing UI based on continued medical user feedback.
- Extending functionality to detect other cancer types or integrate with EHR systems.
This project is dedicated to all the women who continue to fight against breast cancer, to the medical professionals who support them, and to everyone contributing to the global effort to defeat this disease.
MIT License — Feel free to use, modify, and share this tool with attribution.