Imaging activity, not shape
The claim behind positron emission tomography is that a scanner can show what tissue is doing rather than just what it looks like, by tracking a radioactive tracer as it is used up inside the body. The tracer decays by emitting a positron, which travels a short distance before meeting an electron and annihilating with it, producing two gamma rays that fly off in exactly opposite directions. A ring of detectors around the patient catches both rays at almost the same instant, and the line connecting the two detection points tells the scanner where along that line the annihilation happened. Repeated over millions of such events, this builds up a three-dimensional map not of anatomy but of wherever the tracer, and therefore the biological process it is tagged to, was concentrated. That distinction, activity rather than shape, is the whole basis of the method.
Decades of separate pieces
Getting from that physical principle to a usable scanner took decades of separate pieces coming together. Researchers at Massachusetts General Hospital had shown in the 1950s that annihilation radiation could be used for medical imaging at all, and a single-plane scanner built at Brookhaven National Laboratory in 1961 proved the geometry could work outside a physics laboratory. The tracer that made the method clinically useful, a radioactive glucose analogue called FDG, was first synthesised by researchers in Prague in 1968 and was not given to human volunteers until 1976. Michel Ter-Pogossian’s group at Washington University in St Louis supplied the other half of the puzzle: having installed the first cyclotron at a US medical centre in 1963 specifically to produce short-lived isotopes on site, his team, working with Michael Phelps and Edward Hoffman, completed the first multi-slice, cylindrical-array PET scanner in 1974, turning a laboratory demonstration into a repeatable clinical instrument.
What still works exactly as designed
The underlying physics has never been in question, and the engineering built on it has proved durable: coincidence detection of paired gamma rays reliably localises where a tracer has concentrated, and that basic design, refined rather than replaced, is what every PET scanner in use today still does. FDG-based scanning now accounts for the great majority of clinical PET studies, because actively dividing cancer cells take up glucose faster than most healthy tissue and so light up clearly against normal background. Combining a PET scanner with a CT scanner in a single machine, so that metabolic activity can be overlaid directly onto a detailed anatomical image, proved successful enough that a major news magazine named the combined PET-CT scanner its medical invention of the year in 2000, and the fused approach is now the clinical standard rather than a novelty.
The bottleneck the physics built in
What has not been solved is the practical bottleneck built into the physics itself: the isotopes PET depends on mostly have very short half-lives, some measured in a couple of minutes, which means they generally have to be made on site with a cyclotron and used almost immediately. Only fluorine-18, with a half-life long enough to allow commercial production and same-day delivery, has made PET workable outside hospitals large enough to house their own cyclotron, which still leaves most of the world’s hospitals unable to offer it. The images themselves remain noisier than a comparable CT scan, because a PET study is built from a far smaller number of detected photon events, and the geometry can only localise an annihilation event to a region roughly ten centimetres wide before reconstruction software statistically narrows that down, rather than pinpointing it directly.
What it changed in the clinic
The practical effect of being able to image activity rather than shape has been substantial: PET is now central to staging cancer by finding metabolically active tumour tissue that a structural scan alone might miss, to distinguishing types of dementia by mapping where brain glucose use has dropped, and to locating the source of seizures or assessing blood flow to the heart muscle in ways plain anatomical imaging cannot. That benefit comes with a radiation dose that has to be weighed deliberately rather than assumed harmless: a typical scan using the standard FDG dose delivers an effective dose in the same range as some CT examinations, and a combined PET-CT study can deliver several times that, figures clinicians set against the diagnostic information gained before ordering the scan at all.
Is it worth your time
This is worth understanding because it is a rare case where the physics, the chemistry of the tracer and the mechanical engineering of the detector all had to mature roughly in step before the technology became usable, and each piece of that history is traceable to a specific team and a specific year. Readers curious about how a scanner can show a tumour’s activity rather than just its outline will find the mechanism genuinely satisfying once explained. It rewards less anyone hoping for a simple story: the finished device depended on isotopes made minutes before use, tracers first tested on humans nearly a decade after they were synthesised, and a limitation, the short half-life of nearly everything involved, that the field has worked around rather than removed.