Mechanisms of contrast in NMR imaging

Nuclear magnetic resonance pixel intensity and contrast-to-noise has been computed and presented in graphical form for various tissues in the normal central nervous system, on the assumption that the signal intensity is proportional to the macroscopic transverse spin magnetization at the time of detection. T1, T2, and spin density values were experimentally determined using chi-square minimization techniques. Additionally, spin density was derived from partial saturation scans obtained with a long repetition time compared with the spin-lattice relaxation time. Pulse sequences discussed comprise partial saturation, saturation recovery, spin echo, and Carr- Purcell - Meiboom -Gill ( CPMG ). The complicated dependence of signal and contrast-to-noise on the pulse timing parameters and the specific pulse sequence makes it appear desirable to display image intensity so that the dependence on the extrinsic (operator-selectable parameter) is eliminated. Whereas T2 images can be derived from CPMG scans without excessive time penalty, this is not the case for T1 and spin density.