The history
From an Irish physicist's wobbling charges in 1897 to 11.7-tesla images of the living brain.
Physicists spent half a century learning that atomic nuclei are tiny magnets that can be flipped by radio waves. Then, in the 1970s, doctors and chemists asked whether those signals could find disease, and a few inventors worked out how to turn them into pictures. Within a decade MRI was in hospitals, and today it can even watch the brain think.
1938
Magnetic resonance in a molecular beam
Isidor Rabi, New York, USA
1945–46
NMR in ordinary matter
Edward Purcell and Felix Bloch, USA
1950
Spin echo
Erwin Hahn, Illinois, USA
1973
First MR image (two tubes of water)
Paul Lauterbur, New York, USA
1976–77
First image of a living human part (a finger)
Peter Mansfield and Andrew Maudsley, Nottingham, UK
3 July 1977
First whole-body human scan
Raymond Damadian's team, Brooklyn, USA
28 August 1980
First clinically useful patient scan
John Mallard's team, Aberdeen, UK
1986–87
First MRI in India
INMAS, Delhi
1992
First human fMRI
Three US teams
1897 – 1945
Physicists discover that nuclei are tiny magnets that wobble in a field, and that radio waves can flip them.
1897
Studying how magnets change the light from atoms, the Irish physicist Joseph Larmor showed that charges circling in a magnetic field precess: they wobble round the field at a rate set by its strength.
Why it mattered. That wobble, now called Larmor precession, is the frequency every MRI scanner tunes its radio to.
1936
April 1936
Measuring haemoglobin in a magnetic field, Pauling and Coryell found that it is weakly repelled when it carries oxygen, but weakly pulled in when it has given the oxygen up.
Why it mattered. Fifty-four years later this difference became the BOLD signal that lets fMRI watch the brain at work.
1936
1936 and c. 1941
In 1936 the Dutch physicist Cornelis Gorter tried to detect nuclear magnetic resonance in solids and failed. Around 1941 Yevgeny Zavoisky in Kazan saw signals that may have been NMR, but the field of his magnet was not uniform enough to repeat them reliably. In 1944 he discovered the related electron spin resonance instead.
Why it mattered. The effect was real but faint, and finding it needed better magnets and radio electronics.
1938
15 February 1938
At Columbia University, Rabi sent a beam of molecules through magnets and a radio-frequency field. At one exact frequency the nuclei flipped and the beam went astray: magnetic resonance. It won him the 1944 Nobel Prize in Physics.
Why it mattered. It proved nuclei can be flipped by radio waves at a frequency set by the field.
1945 – 1971
NMR is found in ordinary matter and becomes a chemist's tool. Echoes and relaxation times are measured, and tissues turn out to differ.
1946
December 1945 – January 1946
Two teams, working separately, found the signal in everyday materials. Purcell's group at Harvard saw it in paraffin wax on 15 December 1945, and Bloch's group at Stanford saw it in water in January 1946. They shared the 1952 Nobel Prize in Physics.
Why it mattered. For the first time the hydrogen in a lump of matter, or a body, could be made to give a radio signal.
1950
15 November 1950
Hahn hit spins with two short radio pulses and heard an unexpected extra signal afterwards: an echo, as the fanned-out spins came back into step.
Why it mattered. Spin echoes are still at the heart of many MRI scans, and they let scientists measure T2.
1955
After visiting Bloch at Stanford, the Swedish doctor Erik Odeblad built his own NMR equipment and measured human tissues and body fluids. He found that their relaxation times differ.
Why it mattered. It was an early hint that NMR could tell one tissue from another.
1960
March 1960
Soviet records credit Vladislav Ivanov, a young air-force officer, with applying for an inventor's certificate for a way to image the inside of objects with magnetic resonance. It was rejected as unrealistic and only granted in 1984.
Why it mattered. It shows the idea of MR imaging was in the air long before the first pictures.
1971 – 1980
Gradients, slices and Fourier transforms turn NMR signals into images, first of tubes, then of a finger, then of a whole person.
1971
19 March 1971
At the State University of New York Downstate, Damadian reported in Science that tumours in rats had longer NMR relaxation times than healthy tissue. He proposed using NMR to detect cancer and patented a scanning idea, granted in 1974.
Why it mattered. It pointed NMR towards medicine and showed that T1 and T2 could reveal disease.
1973
16 March 1973
Lauterbur had the idea over dinner in 1971: add a gradient so the field, and the frequency, change with position. Nature first rejected his paper, then published it in 1973 with an image of two small tubes of water. He called the method zeugmatography.
Why it mattered. Gradients turned a signal into a picture, and every MRI since uses them.
1974
1973–1974
At the University of Nottingham, Mansfield's group described NMR 'diffraction', the forerunner of k-space, in 1973, and a way to excite just one slice with a gradient in 1974.
Why it mattered. Slice selection lets a scanner image one thin section of the body at a time.
1975
At ETH Zurich, Ernst's group showed how to build MR images with phase and frequency encoding and a Fourier transform. Ernst won the 1991 Nobel Prize in Chemistry for his work on NMR methods.
Why it mattered. Nearly every clinical scanner still fills k-space and uses a Fourier transform this way.
1977
March 1977
Mansfield and Maudsley published cross-sections of a living human finger, made in 1976. The same year Mansfield described echo-planar imaging, a way to make a whole picture from a single pulse in a fraction of a second.
Why it mattered. Echo-planar imaging later made fMRI and diffusion scans possible.
1977
3 July 1977
Damadian's team built a superconducting scanner they named Indomitable. Graduate student Larry Minkoff sat inside for nearly five hours while it gathered 106 points, one at a time, to make a rough cross-section of his chest. The machine is now in the Smithsonian's collection.
Why it mattered. It was the first MR scan of a whole human body.
1980 – 1995
Superconducting scanners go on sale, contrast agents arrive, and MRI spreads around the world.
1980
28 August 1980
The University of Aberdeen's home-built Mark-One scanner, with a 0.04 T magnet, scanned a man with cancer and revealed a tumour in his spine that other tests had not found. It went on to scan thousands of patients.
Why it mattered. It showed MRI could change what doctors did for a patient.
1981
1980–1984
FONAR offered the first commercial whole-body scanner around 1980–81. At Hammersmith Hospital in London, a team from EMI made the first series of brain images in 1981, and GE showed 1.5 T images in 1983. The US FDA approved the first MRI systems for sale in 1984.
Why it mattered. MRI moved from physics labs into hospitals.
1986
1986–87
The first MRI scanner in India was installed at INMAS in Delhi; sources give 1986 or 1987. AIIMS in New Delhi got its first clinical 1.5 T scanner in 1993.
Why it mattered. It began MRI for patients in a country where scanners are now found in cities across the country.
1988
2 June 1988
Gadopentetate dimeglumine, sold as Magnevist, was approved in the United States as the first gadolinium contrast agent for MRI. Injected into a vein, it shortens T1 and makes some tumours and inflamed areas stand out.
Why it mattered. Contrast agents made MRI better at finding disease that plain scans miss.
1985 – 2000
Diffusion and blood-oxygen imaging let MRI follow water in nerve fibres and see which brain areas are working.
1986
Le Bihan showed how to make MR images of diffusion, the random wandering of water molecules. In 1994, with Peter Basser at the US National Institutes of Health, he introduced diffusion tensor imaging, which measures the direction water moves most easily.
Why it mattered. Diffusion scans now spot strokes within minutes and trace the brain’s wiring.
1990
December 1990
At AT&T Bell Laboratories, Ogawa showed in rats that the amount of oxygen in blood changes the MR signal near vessels. He called it blood-oxygen-level-dependent (BOLD) contrast.
Why it mattered. BOLD is the signal behind almost every fMRI study of the working brain.
1992
Three groups published within weeks of each other in 1992, showing the human visual and motor cortex lighting up as people looked at flashing lights or moved their fingers.
Why it mattered. It opened a new science of mapping human thought without surgery or radiation.
2000 – today
Fields climb to 7 T and beyond, safety rules tighten, and low-helium and portable scanners widen access.
2000
c. 2000
Whole-body 3 T scanners were cleared by the US FDA for clinical use between about 1999 and 2002, doubling the field of the common 1.5 T machines.
Why it mattered. Twice the field gives more signal, so sharper or faster scans, especially of the brain.
2001
July 2001
A six-year-old boy having an MRI scan was fatally injured when a steel oxygen cylinder was pulled into the magnet. The American College of Radiology published its first MR safety guidance the next year.
Why it mattered. It is why MRI units now screen everyone and everything that goes near the magnet.
2003
6 October 2003
The Nobel Prize in Physiology or Medicine went to Lauterbur and Mansfield for their discoveries concerning MRI. Damadian was left out. Supporters of Damadian ran full-page newspaper advertisements arguing that he should have shared it. The prize can go to at most three people, and the committee’s discussions stay sealed for 50 years.
Why it mattered. It recognised imaging with gradients, and left a lasting debate about who deserves credit.
2017
12 October 2017
The MAGNETOM Terra became the first 7 T scanner cleared for clinical use: in Europe in August 2017 and by the US FDA on 12 October 2017, for the head and limbs.
Why it mattered. Ultra-high fields reach finer detail, such as tiny lesions in the brain.
2018
11 September 2018
Helium is scarce and its price swings with shortages. Philips' sealed BlueSeal 1.5 T magnet holds about 7 litres instead of about 1,500, with no quench pipe. In 2021 Siemens' 0.55 T MAGNETOM Free.Max was cleared with less than a litre.
Why it mattered. Low-helium magnets make MRI cheaper to run and easier to install.
2020
6 February 2020
The US FDA cleared the first portable MRI scanner, which runs on a 0.064 T permanent magnet and a normal power socket and can be wheeled to a patient's bedside.
Why it mattered. Low-field scanners could bring MRI to places that cannot house a giant magnet.
2023
August 2023 – December 2025
The Bengaluru start-up VoxelGrids unveiled a 1.5 T scanner in 2023 and installed its first clinical system in Chandrapur, Maharashtra, in December 2025. A government project led by SAMEER is building another 1.5 T system for AIIMS Delhi. Both have been called India's first indigenous MRI.
Why it mattered. Home-grown scanners could make MRI far cheaper for Indian hospitals.
2024
2 April 2024
The Iseult scanner, with an 11.7 T magnet, released its first images of the living human brain from 20 volunteers, showing detail 0.2 mm across.
Why it mattered. Stronger fields keep pushing how finely MRI can see the brain.
By the numbers
The magnets used to image people, from Indomitable's weak field to the 11.7 T Iseult. Most hospital scanners today are 1.5 or 3 T.
Did you know?
A 1.5 T scanner's field is about 30,000 times stronger than Earth's, and it stays on day and night, even when nobody is being scanned.
A conventional 1.5 T magnet holds around 1,500 litres of liquid helium. Boiled into gas at room temperature, each litre swells to about 750 litres.
The first whole-body scan, in 1977, took nearly five hours to make one rough picture. Echo-planar imaging can now make a brain image in a fraction of a second.
MRI scanners can be louder than 110 decibels, about as loud as a rock concert, because the gradient coils flex in the magnet's field.
Japan has about 60 MRI scanners per million people, one of the highest rates in the OECD; the UK has about 9.
The people
1857 – 1942 · Physicist · Ireland
Worked out the precession rate that now bears his name.
1898 – 1988 · Physicist · Austria-Hungary and USA
Discovered magnetic resonance in molecular beams; Nobel Prize 1944.
1905 – 1983 · Physicist · Switzerland and USA
Detected nuclear induction in water at Stanford; Nobel Prize 1952.
1912 – 1997 · Physicist · USA
Found NMR in paraffin wax at Harvard; Nobel Prize 1952.
1907 – 1976 · Physicist · USSR
May have glimpsed NMR in 1941 and discovered electron spin resonance in 1944.
1921 – 2016 · Physicist · USA
Discovered the spin echo in 1950.
1936 – 2022 · Physician and inventor · USA
Proposed NMR for cancer detection and built Indomitable, the first whole-body scanner.
1929 – 2007 · Chemist · USA
Made the first MR image using field gradients; Nobel Prize 2003.
1933 – 2017 · Physicist · UK
Invented slice selection and echo-planar imaging; Nobel Prize 2003.
1933 – 2021 · Chemist · Switzerland
Introduced Fourier-transform imaging; Nobel Prize in Chemistry 1991.
1927 – 2021 · Medical physicist · UK
Led the Aberdeen team whose Mark-One scanner made the first clinically useful scan.
born 1934 · Biophysicist · Japan
Discovered BOLD contrast, the basis of fMRI.
Where it happened
Sources
Dates marked “c.” are approximate, and historians sometimes disagree about who was first. If you spot a mistake, tell us.
↑↓ move↵ open/ or ⌘K search anywhere