The full story
This is the episode's narration, word for word. Headings jump to that point in the video.
The question 0:00
How do you make a picture of the inside of a body using nothing but sound? No knife, no X-rays, just a crystal pressed against the skin, listening for echoes. Australia didn't invent that idea. But a small team in Sydney changed what those pictures could show, and a step they championed still runs inside scanners today. To see what they did, we need to start with a shout.
Shouting at a cliff 0:28
Shout at a cliff, and the echo comes back a moment later. The farther the cliff, the longer the wait, because the sound goes there and back. Bats, dolphins and submarines find things this way, with sound above about 20,000 vibrations a second, too high for us to hear. That's ultrasound, and echo time is proportional to distance.
Sonar came first 0:53
During the First World War, that trick became a machine. Paul Langevin and Constantin Chilowsky built a quartz-crystal ultrasonic device to hunt submarines, and it became the basis of naval sonar. Later, some of the first medical experimenters borrowed industrial flaw detectors built on the same idea.
Thirteen microseconds per centimetre 1:16
In soft tissue, sound travels at about 1,540 metres per second. So an echo from one centimetre deep is back in about 13 millionths of a second, and from eight centimetres in about 104. The machine turns each arrival time into a depth, assuming that one speed everywhere. Fat is a little slower and bone far faster, so the depth scale is slightly off wherever tissue isn't average.
Why so high-pitched? 1:47
Why not ordinary sound? Because you can never see detail much smaller than the wavelength you probe with. Wavelength is speed divided by frequency. An audible tone of a thousand vibrations a second has a wavelength in tissue of about one and a half metres. At seven million a second, seven megahertz, it's about a fifth of a millimetre, enough to pick out detail around a millimetre across.
The price of a high pitch 2:17
There's a catch. Tissue soaks up sound as heat, and higher frequencies are soaked up faster. A rough rule says you can see about five hundred wavelengths deep, so seven megahertz reaches roughly 11 centimetres. So it's a trade: high pitch for sharp, shallow pictures like the eye, lower pitch for looking deep into the body. Even so, an echo from ten centimetres down in liver at three and a half megahertz returns over two thousand times weaker in pressure, from the fade alone. So scanners turn up the amplification for later, deeper echoes, a trick called time gain compensation.
The crystal that speaks and listens 3:00
What makes the sound? A crystal. In 1880, Pierre and Jacques Curie found that squeezing certain crystals makes a voltage appear. It works backwards too: apply a voltage, and the crystal changes shape. Inside, it behaves like many tiny electric dipoles; squash it and their charges no longer cancel. That's piezoelectricity. A short voltage pulse makes the crystal ring at ultrasonic frequency, and a returning echo squeezes out a tiny voltage, so one crystal is both speaker and microphone. Modern probes use a ceramic, with a damping layer behind it to keep pulses short.
Why anything inside you echoes 3:46
But why does soft tissue echo at all? Every material has an acoustic impedance: its density times its speed of sound. Wherever impedance changes, some sound bounces back, and the bigger the mismatch, the bigger the echo. From fat to muscle, only about 1.4 per cent reflects, and the rest carries on to make deeper echoes. From tissue to air, about 99.9 per cent bounces back, so there must be no air between probe and skin.
Two kinds of echo 4:19
There are two kinds of echo, and this matters later. Big, smooth boundaries, like the edge of an organ, act like mirrors and send back strong echoes. But structures smaller than the wavelength scatter sound in all directions, and those echoes are much weaker. Together, they make the fine speckled texture that fills the inside of an organ. Those faint whispers are where the information about the inside lives.
From one line to a picture 4:47
Fire one pulse and plot echo strength against depth: that wiggly line is A-mode, for amplitude. Turn each echo into a dot whose brightness stands for its strength, at the depth its timing gives. Step the beam sideways, line after line, and the dots sweep out a slice. That's B-mode, for brightness. Early machines moved the probe on a mechanical arm or through a water tank, with sensors telling the screen where each line belonged.
The pioneers elsewhere 5:21
So who did this first? Not Australia. In Vienna, Karl Dussik published on ultrasound diagnosis from 1942, though his images were later judged to be artefacts. In Denver in 1951, Douglass Howry's team built a two-dimensional scanner inside an immersion tank. In Minnesota, John Wild and John Reid published tissue echo-ranging in 1952. Japanese teams were scanning too. And in Glasgow, Ian Donald, John MacVicar and Tom Brown, who built the machines, published a landmark paper in The Lancet in 1958, which gave convincing evidence that ultrasound was likely to be valuable in obstetrics.
Why a hearing lab looked at babies 6:11
Australia came in through a side door. The Commonwealth Acoustic Laboratories had been set up in 1948 to study hearing. In the late fifties, at the suggestion of a national medical research committee, they employed a physicist, George Kossoff, to look at ultrasound. In 1959, worried about the possible hazards of X-rays in pregnancy, that committee recommended focusing on obstetrics. Doctor William Garrett, a gynaecologist at the Royal Hospital for Women in Sydney, joined as medical consultant.
A bath of water 6:49
David Robinson was hired to build an experimental water-bath scanner for obstetrics, installed at the Royal Hospital for Women in Paddington in 1962. Why water, when commercial machines were hand-held on the skin? Water let them use large transducers with sharper focusing, moved by machinery. And water's impedance is close to tissue's, so less than half a per cent of the sound bounces off where they meet.
An existence proof 7:19
When the first images were shown in America, they were hailed as the equal of any obstetric images, possibly the best in the world, by one insider's later account. Later, by the same account, it was said that the Australian group had provided an 'existence proof': showing what others could do with ultrasound once they caught up.
Outlines only 7:41
But most pictures of the 1960s had a big weakness. Screens were bistable: any echo above a set threshold became a dot of fixed brightness, and anything weaker vanished. You saw edges and shapes, but not echo strength. Even a display tube that could shade its dots handled a range of only about six to one in echo strength, while real echoes span a thousand to one or more. So all those faint whispers from inside organs were thrown away.
Turning up the quiet echoes 8:15
From 1969, the Sydney group, Kossoff, Garrett and engineers George Radovanovich and David Carpenter, attacked that problem. Their insight was that the echoes that matter clinically come from the texture inside soft tissue, not just from its edges. So they designed the electronics to show it. It's called grey scale, and that insider calls it perhaps the group's single most important engineering contribution. An obstetrician whose online history of the field is widely used, Joseph Woo, agrees that true grey scaling grew out of the Kossoff group's work.
How you squeeze a thousand to one 8:57
How do you fit a thousand to one onto a six-to-one screen? You compress it, logarithmically. Log amplifiers already existed, in Denver's scanners for one; Sydney's step was tuning the whole chain to keep the faint texture echoes. Each tenfold step in echo strength gets the same step in brightness, so faint and strong echoes both land on the grey scale. Modern scanners still do it: their data can span over ten thousand to one, far more than a screen of 256 grey levels can show.
Honest credit 9:35
Was it purely Australian? No. In 1965, Siemens' Vidoson, an early real-time scanner, already had basic grey shading. Glasgow's electronics had been grey-scale ready from the start. In London, the Royal Marsden Hospital published grey-scale results in 1973, helped by David Carpenter, a visiting engineer from the Sydney group. Sydney's grey scale began in 1969, before the scan converter, the storage device that let scanners show a grey picture on a TV screen; by 1973 they had built a grey-scale scan converter of their own. Sydney showed grey-scale obstetric scans in Melbourne in 1971, and by 1976 grey-scale machines were widely available.
The Octoson 10:31
The UI Octoson, built in 1975, put eight transducers in a water tank. The patient lay face-down on a flexible membrane over the water and was scanned from below. Its transducers were rings of crystal that could adjust their focus electronically, and fast electronics completed a full compound scan in under a second. The Australian company Ausonics made it, and between 1976 and 1985 more than 200 were manufactured in Australia and sold worldwide.
Overtaken 11:10
Then real-time scanners, with hand-held probes that sweep the beam fast enough for live video, made the Octoson obsolete. In 1989 the Ultrasonics Institute moved into CSIRO. But squeezing faint and loud echoes onto one grey scale is still a standard step in scanners today. Whether you ever need a scan is a question for a doctor; this video is only the physics and the history.
The next question 11:41
Next time, a different mystery, this one inside the stomach. In 1985, Barry Marshall and colleagues published a paper with a blunt title: an attempt to fulfil Koch's postulates for a stomach bacterium. So how do you prove that a germ causes a disease?
Sources
Every factual claim in the episode is tied to one of these. Spotted an error? Tell us.
- OpenStax, College Physics 2e, §17.7 "Ultrasound" (CC BY 4.0) — openstax.org
- J. Woo, "A short History of the development of Ultrasound in Obstetrics and… — ob-ultrasound.net
- R. Gill, "Medical ultrasound in Australia: A short history", Australasian Journal of… — pmc.ncbi.nlm.nih.gov
- J. Woo, "A short History of the development of Ultrasound in Obstetrics and… — ob-ultrasound.net
- J. C. Lacefield, "Ultrasound Imaging", Chapter 13 in D. R. Dance et al. (eds),… — www-pub.iaea.org
- J. C. Lacefield, "Physics of Ultrasound", Chapter 12 in D. R. Dance et al. (eds),… — www-pub.iaea.org
- J. J. Wild and J. M. Reid, "Application of Echo-Ranging Techniques to the… — doi.org
- I. Donald, J. MacVicar and T. G. Brown, "Investigation of Abdominal Masses by Pulsed… — doi.org
- K. J. W. Taylor, D. A. Carpenter and V. R. McCready, "Grey scale echography in the… — doi.org
- B. J. Marshall, J. A. Armstrong, D. B. McGechie and R. J. Clancy, "Attempt to fulfil… — doi.org
Researched and scripted with AI assistance, fact-checked claim by claim, with synthetic narration and diagrams drawn in code. How we make episodes.
