🩻 2D CT Scanner: Sinogram & Filtered Back-Projection

🔬 What It Demonstrates

A simulated X-ray beam sweeps 0-180° around a synthetic tissue cross-section (the "phantom"). At each angle the simulator computes a genuine Radon transform — a rotated line-sum of the density map — and stacks that 1D projection into a growing sinogram. Every projection is then back-projected to rebuild the slice; with the ramp filter on, each projection is convolved with a discrete Ram-Lak kernel before back-projection, which is the same filtered back-projection (FBP) math real CT scanners use to turn a blurry sum into a sharp image.

How to Use

• Pick a Phantom shape and watch the sinogram build row by row as the beam angle advances.
• Projection angles sets how many angles the 180° sweep is sampled at — fewer angles leave visible streak artifacts in the reconstruction, exactly as under-sampling does on a real scanner.
• Toggle Ramp filter (FBP) off to see plain (unfiltered) back-projection produce a blurred reconstruction, then on to see the sharpened filtered result.

Did You Know?

A single point of dense tissue traces a sine-wave curve across a real sinogram as the scanner rotates around it — that's literally where the word "sinogram" comes from. Unfiltered back-projection always produces a characteristic 1/r blur, which is exactly why the ramp filter exists in real scanners.