My Journey from WebDev to Medical Visualization Rustacean
summary
Builds a working Rust pipeline that reads DICOM stacks (about 295 frames in the example), decodes Hounsfield values, and displays them with a raycasting volume renderer that supports camera orbit, threshold UI and three selectable shaders. Implementation moves from three-d's example to custom GL/glow shaders then is ported to wgpu as a proof of concept. Demonstrates three raycasting shaders (default, MIP, average projection), explains sampling choices and shows which approach yields the clearest clinical detail.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:03From webdev to medical visualization
- 2:29Presentation overview
- 3:54Learning Rust
- 5:45Medical visualization graphics
- 7:10Rendering libraries and crates
- 8:14DICOM and medical data
- 10:05First DICOM volume renderer
- 11:10Raycasting for volume rendering
- 13:19Rendering challenges and fixes
- 14:07Live raycasting demo
- 15:12Rendering libraries comparison
- 16:39Porting to wgpu
- 17:19Library evaluation
- 18:48Closing thoughts
related talks
38mTokio, Rayon, and the GPU: Profiling a Radiology AI Pipeline in Rust
02h 03m2D Graph scene in Rust
02h 13m2D Graph scene in Rust
24mArtisanal Rendering: Make Your Own Computer Art!
31mSplit Rendering TV Applications Using WASM
23mCraig's Angular/Rust Spectacular
26mThe art of programmatic videos with Rust
32mData science in the browser: DX & UX