IP / DVR

CIBC

Image ProcessingDirect Volume Rendering

IP / DVR

CIBC

Image Processing in SCIRunFour primary options:

• Native SCIRun– Interpolation– Gradient– TransformFieldData

• Teem (Nrrd, Gage, Tend, …)– N-dim raster data “Swiss Army Knife”– Crop, slice, permute– Local measures (via Gage and Tend)

• ITK– Similar filtering operations to Teem– Segmentation filters (threshold, confidence-connected, level

sets, …)– Registration

• MATLAB

IP / DVR

CIBC

TeemLearn “unu” and “tend” (Verbs of

Raster Processing)Decompose Complex Tasks Into

Simple StepsAccurate Kernels

• Derivatives between sample points

IP / DVR

CIBC

Teem in SCIRun

IP / DVR

CIBC

ITK in SCIRun

IP / DVR

CIBC

Direct Volume Rendering

Multi-dimensional Transfer Functions• Boundaries• BioImage

More ExamplesSCIRun Volume Rendering Modules

IP / DVR

CIBC

Gordon Kindlmann’s MS Thesis

IP / DVR

CIBC

Boundaries

IP / DVR

CIBC

Boundaries

IP / DVR

CIBC

Boundaries

IP / DVR

CIBC

Boundaries

IP / DVR

CIBC

Boundaries

IP / DVR

CIBC

Tumor Vessel Imaging and Visualization

+ + +

Immobilize.Dose Contrast

OptimizeSignal : Noise vs Time

Discern Contrast Detect Boundaries

G. Kindlmann, D. Weinstein, G. Jones, C.R. Johnson, M. Capecchi, and C. Keller. Practical Vessel Imaging by Computed Tomography in Live Transgenic Mouse Models for Human Tumors, Journal of Molecular Imaging, 2005.

IP / DVR

CIBC

Volume Rendering: SIMIAN (Joe Kniss)

QuickTime™ and aYUV420 codec decompressor

are needed to see this picture.

IP / DVR

CIBC

Mouse MRI – Al Johnson - Duke

IP / DVR

CIBC

Volume Rendering in SCIRunTexture ObjectsGradientsSlice Rendering

• Axis aligned• Tangent to view direction

Volume Rendering• Slice based• MIP (max operator)• “Direct volume rendering” (over operator)

IP / DVR

CIBC

Volume Rendering in SCIRun

IP / DVR

CIBC

For More Information:http://www.sci.utah.eduhttp://www.sci.utah.edu

{dmw,jeroen,dav}@sci.utah.edu{dmw,jeroen,dav}@sci.utah.edu

IP / DVR

CIBC