| John Lai 2005-07-03, 12:33 pm |
| Hello George,
Thanks for your informative insight. Very interesting demo at your site.
John
<buyanovsky@attbi.com> wrote in message
news:1120203200.749867.211410@g44g2000cwa.googlegroups.com...
> Hello John,
>
>
> It is not a proprietary; we use tri-linear interpolation.
>
>
> The tri-linear interpolation is totally legitimate to use in the
> medical applications. It preserves the data range; means that
> interpolated values remain within range of interpolating nodes. Matter
> of fact, an oblique MPR of all professional medical workstation uses
> tri-linear interpolation. The same is true for volume rendering - all
> professional medical workstation uses tri-linear interpolation. The
> high quality rendering of our solution is the result of supersampling
> - density of sampling points along ray is 32/16 sampling points per
> voxel/cell. Each sampling point is the output of tri-linear
> interpolation. Each pixel on projection plane is source of ray so;
> density of rays depends on zoom (for zoom 16 - the density is 16 rays
> per voxel).
>
>
> Curious statement. All images you can see
> on our web site <www.fovia.com> are rendered
> from CT/MRI data sets 1-5 years old. The size
> of data sets are within range 512 x 512 x [300...1800].
>
> Regards,
> George
>
>
> John Lai wrote:
>
|