Mansfield's echo-planar scanning grabbed a whole MRI image in a fraction of a second

Magnetic resonance imaging emerged in the early 1970s when Paul Lauterbur showed that gradient magnets could turn nuclear spin signals into two-dimensional images. But early scanning strategies acquired spatial information a line at a time, so a useful image took minutes — and a living, breathing patient moves far too much for that. Speed was not a convenience; it was the difference between MRI as a lab curiosity and MRI as medicine.

Peter Mansfield's group at Nottingham attacked the repetition itself. Their 1977 multi-planar and echo-planar methods, published in Journal of Physics C as 'Multi-planar image formation using NMR spin echoes', used rapidly varying gradients to collect enough signal for a complete image plane after a single excitation — echo-planar scanning, as it came to be called. The Nobel Assembly later credited Mansfield precisely with showing 'how extremely rapid imaging could be achieved by very fast gradient variations.'

The technique outran its own hardware: gradient coils and amplifiers took about a decade to become clinical-grade, but once they did, echo-planar imaging became the workhorse of the fastest applications — cardiac imaging, stroke diffusion imaging, and essentially all of functional MRI.

Encode the whole plane at once, not point by point

A one-shot acquisition freezes physiology: motion artefacts shrink to nothing when the whole image is captured in tens of milliseconds instead of minutes.

The gradients do the encoding work instead of repeated excitations, so speed came from better engineering of a switchable magnet rather than from waiting for stronger magnets.

Mansfield's signal analysis made the rapid raw echoes mathematically reconstructable, turning a physics curiosity into 'a practical method', as the Nobel citation put it.

The approach anticipated the hardware by a decade — publishing the method early meant the clinical systems that matured in the late 1980s adopted it wholesale.

When throughput is the bottleneck, look for the step that repeats needlessly. Encoding everything in one pass trades hard engineering on the switch hardware for an order of magnitude in speed.

The Nobel Prize in Physiology or Medicine 2003 went jointly to Lauterbur and Mansfield 'for their discoveries concerning magnetic resonance imaging.' By then echo-planar imaging underpinned virtually all fMRI and the diffusion imaging used to map the brain's fiber tracts, per the 2010 PLOS review literature — and whole-brain fMRI still runs on it, with later refinements like multiplexed EPI tackling the remaining seconds-long multi-slice acquisition. A method first written down in 1977 became the standard way neuroscience watches the brain think.