📚 Syllabus & Course Curriculum
AI & Machine Learning in Healthcare
Module-by-module breakdown of Deep Learning for Histopathology: WSIs, MIL & Transformers, from foundations to a certified capstone project.
Slides
• WSI formats, pyramidal storage and reading tiles efficiently
• Stain variation across scanners and laboratories, and normalisation methods
• Tissue detection, artefact rejection and quality control at scale
Patches
• Magnification selection and the context-versus-detail trade-off
• Augmentation appropriate to histology, including stain augmentation
• Self-supervised pretraining on unlabelled slides
MIL
• The weak-label problem: one diagnosis, thousands of tiles
• Attention-based MIL and instance aggregation strategies
• Interpreting attention heatmaps without overclaiming localisation
Transformers
• Vision transformers on histology tiles
• Modelling spatial relationships between regions rather than tiles in isolation
• Foundation models for pathology and how to evaluate their transfer
Clinical
• Multi-site validation and scanner generalisation failure
• Regulatory expectations for computational pathology tools
• Integration with the laboratory information system and pathologist workflow
Earn government-registered certification in Deep Learning for Histopathology: WSIs, MIL & Transformers
e-Certificate and e-Marksheet issued on successful completion.
