1 / 19

A Plan for Visualization and Mapping Neurons

A Plan for Visualization and Mapping Neurons. Ying Zhu Department of Computer Science Georgia State University. Project Goal. Create a intuitive web-based mechanism for recording and visualizing on a 3-D brain model the locations of: identified neurons, observed neurons,

ponce
Download Presentation

A Plan for Visualization and Mapping Neurons

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. A Plan for Visualization and Mapping Neurons Ying Zhu Department of Computer Science Georgia State University

  2. Project Goal Create a intuitive web-based mechanism for recording and visualizing on a 3-D brain model the locations of: • identified neurons, • observed neurons, • neuron arborization.

  3. User Interface • Create a 3D brain atlas • Allow users to manually deform the 3D atlas to match the target brain • Can overlay brain images to guide the deformation • Users can mark the location of the neurons on the deformed 3D model • Enter additional neuron attribute information.

  4. Creating the Model • We create the 3D model by breaking down the brain shape into 4 regions. • Connecting the lobes will happen later. • Nerves will also be added later 2 3 4 1

  5. Deform the Brain Model • Physics based deformations are difficult to control. • We use non-physics based deformation algorithms • More efficient • Users have interactive control over the deformation process

  6. Select the Location • Using a mouse or virtual laser pointer, the user selects a point that they believe represents the location of the neuron.

  7. How to represent neuron locations? • We want to conduct comparative studies of homologous neurons • But brains come in different sizes and shapes • To compare their locations we need a common coordinate system that adapts to shape/size variations

  8. Existing Methods • Create a 2D or 3D brain atlas from brain cross section images • Define a coordinate system for the atlas based on certain landmarks • Deform the atlas to fit the target data (images), or vice versa • Eventually assign a coordinate to the features on the target images

  9. Characteristics of Tritonia Brain • Most neurons are on the surface or just below the surface of the brain • Cross section images are not necessary as in the study of other types of brains. • We want to mark the neuron locations directly on a 3D brain model • The shape of Tritonia brain is relatively simple • But we still need to deal with shape and size variations

  10. Our Solution • Use “texture mapping” technique to establish a direct mapping between the 3D model and a 2D image • Each point on the 3D brain model is mapped to a point on the 2D image

  11. What is Texture Mapping? • Texture mapping is the method of taking a flat 2D image of what an object's surface looks like, and then applying that flat image to a 3D computer generated object. • Makes a surface look textured even though geometrically it isn’t.

  12. Texture Mapping • Using the texture mapping technique we can map a (x,y,z) value a (s, t) value, or vice versa. • Forwardx = X(s,t)y = Y(s,t)z = Z(s,t • Reverses = S(x,y,z)t = T(x,y,z) • And… we can deform the object, but the (s,t) texture for that vertex does not change. (x,y,z) t s

  13. Texture Mapping Example • Each pixel on the 2D image is assigned to a vertex on the 3D object.

  14. Neuron Mapping • The 3D neuron location is mapped to the 2D “texture” image coordinate • The 2D coordinate will be stored in database • Can easily project this 2D coordinate back to the 3D brain atlas • The size and shape of the brain model may change but the mapping between 3D model and 2D image is stable • The 2D “texture” image thus provides a common coordinate space

  15. What is stored? • Final data stored for Neuron location is: • Which region (lobe) the point was marked. • The 2D texture coordinates (s,t) of the point that was marked. Coord: 2 3 3 4 1 , 150,380

  16. Benefits • Does not rely on brain cross section images • User-guided, semi-automatic model deformation for better control • Allow user to mark neuron locations directly on 3D models • Address brain shape/size variations by mapping neuron locations to a common 2D coordinate space • Can be easily adapted to different species • Use different 3D atlas, “texture” image, and/or mapping equations

  17. Limitations • Only works for brain models where neurons are on or close to the surface • It will take some practice to learn how to deform the atlas model. • Texture Mapping is not a one-to-one mapping, so some inaccuracies may result. • Can be minimized by matching the resolution of the 2D image with that of the 3D atlas • How accurate do we need to be?

  18. Acknowledgements • People • Paul Katz (Biology) • Raj Sunderraman (CS) • Jason Pamplin (CS) • Robert Calin-Jageman (Biology) • Jim Newcomb (Biology) • Hao Tian (CS, Brains & Behavior Fellow) • Christopher Gardner (CS, Brains & Behavior Assistantship) • Lei Li (CIS, Brains & Behavior Fellow) • Brains & Behavior Program

  19. Questions to Neuroscientists • Do you foresee any problem with this approach (particularly the neuron localization)? • Will this brain mapping approach be useful for brains of other invertebrate species? • How do you save neuron location information in your database? • Can there be a general way to specify neuron locations for invertebrates?

More Related