Skip to content

Biomedical Imaging — Overview

Notes on biomedical imaging: the physics of imaging modalities and the math of reconstructing images from measurements.

Topics to cover

  • MRI: spin physics, k-space, pulse sequences
  • CT: X-ray attenuation and the Radon transform
  • Ultrasound and optical imaging
  • Reconstruction, denoising, and deep-learning approaches

CT and the Radon transform

CT reconstruction inverts the Radon transform, the set of line integrals of the attenuation f(x,y)f(x, y) through the body:

Rf(θ,s)=f(scosθtsinθ,  ssinθ+tcosθ)dtR f(\theta, s) = \int_{-\infty}^{\infty} f\big(s\cos\theta - t\sin\theta,\; s\sin\theta + t\cos\theta\big)\, dt

X-ray attenuation along a path follows the Beer–Lambert law:

I=I0exp ⁣(μ(x)dx)I = I_0 \exp\!\left( -\int \mu(x)\, dx \right)

where μ(x)\mu(x) is the linear attenuation coefficient.

MRI signal

The measured MRI signal is the Fourier transform of the transverse magnetization, sampled in k-space:

s(k)=m(r)ei2πkrdr.s(\mathbf{k}) = \int m(\mathbf{r})\, e^{-i 2\pi \mathbf{k}\cdot\mathbf{r}}\, d\mathbf{r}.

Next steps

Add pages such as mri.md, ct.md, or reconstruction.md to this folder.