Difference between revisions of "Basic Terminologies"

From CRIPTWiki
Jump to: navigation, search
Line 39: Line 39:
  
 
So far, we have not dealt with any fMRI imaging data but I imagine with 4D data processing, fMRI should not be too huge a concern.  
 
So far, we have not dealt with any fMRI imaging data but I imagine with 4D data processing, fMRI should not be too huge a concern.  
 +
 +
===Segmentation===
 +
The process of partitioning a digital image into multiple regions (sets of pixels).
  
 
===Surface Generation===
 
===Surface Generation===
 +
 
===Volumetric Rendering===
 
===Volumetric Rendering===
 +
 
===Isosurface===
 
===Isosurface===
 +
a three-dimensional analog of an isocontour. It is a surface that represents points of a constant value
  
 
[[Category:Amira]]
 
[[Category:Amira]]

Revision as of 15:09, 14 May 2009

Pixel

  • A square display unit. All 2D pictures are composed of pixels.
  • A 3 pixel by 3 pixel pictures contains 9 pixels.
  • Pixel is flat with unit length of 1 and width of 1.

Voxel

  • A cube display unit.All 3D data sets are made from voxels.
  • A 3 voxel by 3 voxel by 3 voxel matrix contains 27 voxels.
  • voxel is 3D and has length of 1 and width of 1 and height of 1.

2D Imaging

eg. Histology Data measurment including length and width.

  • NOTE: histology is always considered 2D but it does include height. Hence if you are given many histological slices, you can reconstitute them into a 3D data set as well.

3D Imaging

eg. Modeling, plastic, virtual,etc Data measurement including height, length and width.

  • Note: Model is considered 3D most of the time. However, once you start dynamically swapping 3D data sets (eg. a heart beating through time). Hence, a series of 3D data sets of relevant structures are considered 4D.

4D Imaging

Data measurement including height, length, width and time. Time is the key here. This is the hardest dataset to work with. Make sure you recycle labels should it be necessary.

CT/CAT

  • Known as Computed Tomography (CT)or Computed Axial Tomography (CAT)
  • The basic idea behind CT is the idea of density.
    • In CT, the denser the object, the more bright it appears in the final result.
    • Hence a tissue of heterogeneous density creates a gray scale spectrum. Each voxels will have its own gray scale value (each voxel can only have one value only just like how pixel cannot be two colors.)
  • CT is in many sense very similar to advanced array of X ray machines.
    • Traditional X ray is only exposed ONCE (eg. you stand in front of the X ray machine and get shoot via X rays for once and develop the filem.) However, CT has more exposures and at lower dosages during each exposure. Simplest way I can explain it would be try to imagine a Circular track with an X ray emitter mounted on the track and the patient inside the circular ring while the receiver at the opposite end of the circular track. As the scan proceeds, the emitter starts at 12 o clock position with the receiver at 6 o clock position, the density reading from 12 to 6 position is ready... and then the emitter/detector moves along the track and keeps take reading... The ending result is a 2D reading of the plane being scanned(This is EXTREMELY simplified explanation). Here is a BRIEF intro to CT/CAT principles.

MRI

Nuclear Magnetic Resonance Imaging They take out "Nuclear" because they don't want to scare the public. It is must less intuitive to grasp the understanding of MRI. I do not believe I am qualified to really explain it well so I will share my view with you. basicly, a HUGE magnet is used to generate a radiofrequency pulse that excites particles (1H or 13C) inside the issue and once the magnetic field is reduced, the energy absorbed by the excited particle is re-emitted back to the surroundings and picked up by the MRI.

The signal strength is not significantly influenced by the density of the object but more by the concentration of the excitable particles and the way radio pulse was generated (echo time and repetition time). I highly recommend at least skim through the Wikipedia articles on MRI. I think the most important information you will need is to be able to differentiate T1, T2 weighting and what they each excel at visualizing.

So far, we have not dealt with any fMRI imaging data but I imagine with 4D data processing, fMRI should not be too huge a concern.

Segmentation

The process of partitioning a digital image into multiple regions (sets of pixels).

Surface Generation

Volumetric Rendering

Isosurface

a three-dimensional analog of an isocontour. It is a surface that represents points of a constant valueFile:Example.jpg