The Wellcome Foundation lecture, 1984. Nuclear magnetic resonance imaging in medicine: medical and biological applications and problems.

From early biological work and the first T1 nuclear magnetic resonance (n.m.r.) animal image in 1974, whole-body patient images, by using a two-dimensional Fourier transform method were achieved in Aberdeen in 1980 with a 0.04 T vertical resistive magnet. Different pulse sequences produce images dependent by different amounts on proton density, T1 and T2, and for clinical work it is advantageous to use more than one pulse sequence to image pathology. The slow improvement of spatial resolution with increasing standing magnetic field strength is discussed and information on the T1 and T2 contrast dependence is reviewed: it suggests that the gains from high fields may be less than believed hitherto. Electrocardiogram gating can be used to produce moving images of the beating heart; blood flow can be imaged and surface radiofrequency coils are used for improved detail. N.m.r. imaging has considerable potential for studying response to therapy; mental states and dementia; tissue generation; discriminating body fat and body fluids. Other nuclei such as 23Na can be imaged and the potential to image fluorine-labelled pharmaceuticals could be very exciting; n.m.r. contrast agents are now being developed. Images formed from T1 values measured for each pixel are very useful for diagnosis, but the numerical values themselves are less valuable for distinctive pathological identification. With 15 companies manufacturing n.m.r. imagers and over 200 in use in hospitals, the technique is rapidly becoming established in diagnostic clinical practice and some typical uses are presented.