3D Surface vs Volume Rendering

OsiriX and Horos Course · 3D Reconstruction

In this lesson I explain the difference between the two main types of 3D reconstruction in OsiriX: 3D Volume Rendering and 3D Surface Rendering. Although both generate 3D images, the working principle, the type of result and the intended use are quite different. 1) Concept: pixels/voxels (Volume) vs vector (Surface) Volume Rendering (voxel-based) The CT scan generates a volume made up of "little dots": pixels (2D) and voxels (3D) Each voxel has characteristics such as density. In Volume Rendering, OsiriX uses these voxels and lets you apply: color opacity/transparency density-based adjustments Result: a very "malleable" 3D image for evaluation, where you can progressively reveal layers (skin, subcutaneous tissue, veins, muscles, bones, etc.) just by adjusting transparency/contrast. Surface Rendering (vector / mathematical image) In Surface Rendering, OsiriX "transforms" the volume into a vector representation (a mathematical description). You define a density threshold: OsiriX creates a surface based on that cutoff. The result tends to be excellent for: exporting using in 3D software modeling / 3D printing And an important characteristic shown in the video: when you zoom in, the surface does not depend on pixels like Volume Rendering does — it is drawn by mathematical lines/shapes. 2) Example in OsiriX: Volume Rendering I use a simple lower-limb case just to demonstrate the principle. I open 3D Volume Rendering I crop to reduce the block I show the impact of changing contrast/transparency: less dense structures become transparent first progressively revealing deeper layers This is the great strength of Volume Rendering: exploring and evaluating the volume with quick adjustments. 3) Example in OsiriX: Surface Rendering I open 3D Surface Rendering I choose to use just one surface (more didactic) I define a density cutoff point (threshold) In the example, with this threshold it extracts the bones (the denser structures) well Comment from the video: you can adjust the settings to get more "perfect" surfaces, depending on the goal. 4) Analogy with Photoshop and Illustrator (to understand the difference) In the video I make a comparison: Photoshop: image made of pixels if you zoom in a lot, you start to see the "little squares/dots" you can adjust brightness/contrast with great flexibility Illustrator: vector image you "trace" the image and it becomes mathematical lines/shapes when you zoom in, you don't see pixels — you see mathematically defined curves/lines This logic is similar to: Volume Rendering ↔ "pixel/voxel" (more malleable for visualization) Surface Rendering ↔ "vector" (better for exporting and working in external 3D software) 5) When to use each one (as in the video) Volume Rendering Best for evaluation and dynamic manipulation of the image: adjusting contrast/opacity quickly exploring different structures Surface Rendering Best when you need a "mathematical" 3D object for: exporting editing in 3D software applications such as modeling (e.g., prosthesis / planning) and external manipulation 6) Exporting the Surface (OBJ) and opening it in 3D software At the end I show: Exporting the Surface in a common format (e.g., OBJ) Opening that file in a (non-medical) 3D application to demonstrate that: you can apply colors edit manipulate the mesh and create different effects/visualizations The goal is to make the difference between the two 3D reconstruction modes in OsiriX clear.