k-Space in MRI
What is k-Space?
k-space is the frequency domain representation of an image, mainly used in MRI (Magnetic Resonance Imaging).
- Stores frequency and phase information
- Not directly viewable like a normal image
Simple Definition
k-space = raw data collected by MRI scanners before image reconstruction
How it Works (Step-by-Step)
- MRI scanner collects signals from the body
- Data is stored in k-space (frequency domain)
- Apply inverse Fourier Transform
- Final image (spatial domain) is obtained
- Normal image → shows position (x, y)
- k-space → shows frequency components
What Different Parts of k-Space Mean
| k-space Region | Meaning |
|---|---|
| Center | Low frequency → overall shape and brightness |
| Edges | High frequency → edges and fine details |
An image is made of many waves. k-space stores those wave components.
Important for Machine Learning & Radiomics
- DICOM images are already reconstructed
- You do NOT work with k-space in most ML tasks
img = dcm.pixel_array # spatial domain image
Why k-Space Matters
- Faster MRI scanning
- Image reconstruction research
- Deep learning for reconstruction
k-space is the frequency-domain representation of MRI data. It contains raw signal information that is transformed into a spatial image using an inverse Fourier transform.
k-space = frequency domain → DICOM = spatial image