Medical Imaging And 3-d Computer Graphics

3G4 Medical Imaging and 3D Computer Graphics

timing: Michaelmas term, 16 lectures
assessment: written exam (1.5 hours) at the start of Easter term

If you are considering taking this course, and want a rough idea of what it involves, click on the resources below.

The Laboratory2010 Tripos ExaminationSample Examples Paper
skull.gifct.gifshrek.jpg
external link: 3G4 laboratory handout
external link: 3G4 lab sign-upexternal link: questions and external link: solutionsexternal link: 3D graphical rendering examples paper

This module aims to introduce state-of-the-art techniques for the acquisition, representation and visualisation of structured 3D data. The main application area considered in the module is diagnostic medical imaging: 3D data is acquired using one of the popular imaging modalities (e.g. CT), represented as a voxel array or segmented into surfaces, then visualised using advanced computer graphic techniques. While medical imaging is the focus of the course, many of the techniques used to segment, represent and visualise the 3D data sets are generic and can equally be applied to other types of data, such as CAD models.

On completion of the module, students should:

  • Be able to explain the principles of operation of CT, nuclear medicine and diagnostic ultrasound and magnetic resonance imaging;
  • Be aware of the advantages and risks associated with these techniques and understand the types of diagnostic problems that each can address;
  • Be aware of other types of data to which segmentation and visualisation algorithms can be applied (eg. CAD models);
  • Understand the different ways to represent 3D data and appreciate the advantages and disadvantages of each technique;
  • Know how to extract surfaces from volumetric data;
  • Be aware of the range of computer graphics algorithms and hardware used to visualise 3D data;
  • Understand how surfaces can be rendered using suitable illumination and reflection models;
  • Know how to visualise voxel arrays directly using volume rendering techniques.

Leader: external link: Dr. Andrew Gee
Additional lecturers: external link: Prof. Richard Prager, external link: Dr. Graham Treece

Online Course Material

external link: 3G4 laboratory handout

external link: 3G4 lab sign-up

external link: Curves, surfaces, interpolation and rendering demos

Topics

Medical Image Acquisition (5L, Prof. R. Prager)

  • X-rays and the Radon transform
  • Tomographic reconstruction algorithms in both the spatial and frequency domains
  • Emission computed tomography: SPECT and PET
  • Iterative reconstruction algorithms
  • 2D and 3D ultrasound
  • Introduction to Magnetic Resonance Imaging

Extracting Information from 3D Data (6L, Dr G.M. Treece)

Polygonal representations and efficient storage

  • Parametric curves and surfaces
  • Subdivision and display of parametric surfaces

Surfaces from sampled data

  • Thresholding, morphological operators and contours
  • Surface extraction - marching cubes

Interpolating sampled data

  • Interpolation of isotropic data
  • Distance transforms and interpolation of non-isotropic data
  • Unstructured data - RBFs and Delaunay triangulation

Direct surface capture

  • Laser stripe scanners
  • Space encoding: the cubicscope

3D Graphical Rendering (5L, Dr A. Gee)

  • Viewing systems: viewpoints and projection
  • Reflection and illumination models: the Phong reflection model
  • Surface rendering: incremental shading techniques, hidden surface removal using Z-buffers
  • Shadows and textures
  • Ray tracing
  • Volume rendering
  • Computer graphics hardware

References

see also external link: Booklist for IIA courses