Medical Image Classification Pipeline with Deep Learning

Create a comprehensive Python workflow using TensorFlow and Keras to develop a multi-class medical image classification system for detecting potential lung abnormalities from chest X-ray datasets. The pipeline must include automated data preprocessing, augmentation strategies, model training with transfer learning, and a performance evaluation framework that calculates precision, recall, and F1 score across different disease categories. Implement robust cross-validation techniques and generate a detailed performance report with confusion matrix visualization.