Medical Image Classification Convolutional Neural Network

Develop a deep learning model using TensorFlow and Keras for automated medical image classification across multiple diagnostic domains (radiology, pathology, dermatology). The system must support transfer learning, handle multi-class and multi-label classification scenarios, and provide uncertainty quantification. Implement data augmentation strategies specifically tailored to medical imaging, with support for handling imbalanced datasets. Create a modular architecture allowing easy retraining and integration of new image classification tasks.