Magnetization, resonance and relaxation
MRI begins with nuclear magnetization in a strong static magnetic field. RF energy perturbs that system, and the return toward equilibrium is described through longitudinal T1 recovery and transverse T2 decay. Sequence timing samples those processes differently, creating contrast between tissues.
Gradients and spatial encoding
Gradient fields deliberately vary magnetic field strength with position. Slice selection, phase encoding and frequency encoding use those controlled variations to localize signal. Understanding gradients makes it easier to reason through resolution, field of view, aliasing, echo formation and many artifacts.
What k-space represents
K-space is the raw spatial-frequency data matrix used to reconstruct an MR image. Central k-space contributes strongly to broad image contrast and signal characteristics, while outer regions contain higher spatial-frequency detail. Acquisition trajectory, ordering and undersampling strategies affect how information is collected and how artifacts appear.
SNR, resolution and scan time
MRI optimization is a tradeoff problem. Larger voxels generally collect more signal, while higher spatial resolution usually reduces voxel volume and SNR. More averages can improve SNR but increase scan time. Receiver bandwidth, matrix, FOV, slice thickness and acceleration choices interact, so changes should be evaluated as a system rather than in isolation.
From physics to troubleshooting
Physics becomes useful when it predicts the image. Motion, susceptibility, chemical shift, wrap, flow effects and noise all have physical causes. A solid foundation lets technologists choose the right corrective direction instead of relying only on memorized fixes.