<computer_vision for |>

Module 1: Python and image data essentials

A focused primer on the Python you need to load, manipulate and visualise medical images. Skip it if you already code.

Module 2: Foundational mathematics for vision

The maths you actually need — vectors and matrices, convolutions, probability, and the intuition behind gradient descent and backpropagation. Taught for clinicians, no PhD required.

Module 3: How machines see images

Pixels, tensors, colour channels, and how a neural network turns raw imagery into predictions. The mental model everything else builds on.

Module 4: Convolutional neural networks for radiology

Build and train CNNs on X-ray and CT data. Understand what convolutional layers are actually learning.

Module 5: Segmentation for radiology (CT and MRI)

U-Net and modern segmentation architectures for outlining anatomy and pathology. Where these models are already in clinical use.

Module 6: Whole-slide imaging for histopathology

How pathology AI handles gigapixel slides, tile-based inference, and the practical constraints of digital pathology workflows.

Module 7: Video and endoscopy

Applying vision models to endoscopy and other clinical video. Temporal considerations, frame-level vs sequence-level models.

Module 8: Bias, fairness and safety in imaging AI

How imaging models fail in the real world, where bias enters, and what to look for before trusting a model on your patients.

Module 9: Critical appraisal of imaging AI

A framework for reading imaging AI papers and evaluating commercial and FDA-approved tools with clinical rigour.

Module 10: Capstone: your clinical image classifier

Ship a portfolio-ready notebook end-to-end. Something you can share, reference at interview, or extend into your own project.