Disease Classification Using CNN and Feature Fusion

A state-of-the-art machine learning project for automated skin lesion classification. This system leverages Convolutional Neural Networks (CNN) combined with patient metadata feature fusion to provide accurate predictions of six common skin conditions: Actinic Keratosis (ACK), Basal Cell Carcinoma (BCC), Melanoma (MEL), Nevus (NEV), Squamous Cell Carcinoma (SCC), and Seborrheic Keratosis (SEK).

Key Highlights:

- Hybrid Approach: Combines deep learning for image analysis with ensemble learning (XGBoost, LightGBM, Random Forest) for robust predictions.

- Feature Fusion: Integrates CNN-extracted image features with patient demographics, medical history, lesion characteristics, and environmental factors.

- Interactive Web App: Built with Streamlit, allows image upload, multi-step questionnaires, and real-time disease predictions with confidence scores.

- Comprehensive Functionality: Patient profile management, progress tracking, and ensemble-based decision support for medical professionals.

Technologies Used: Python, TensorFlow/Keras, Streamlit, Scikit-learn, XGBoost, LightGBM, Pandas, NumPy, OpenCV.

Usage: Upload skin lesion images, answer patient history questions, and receive instant predictions to assist in clinical decision-making.