Godfrey Hounsfield’s groundbreaking concept of viewing organs from outside the body was so ambitious that it would require the most successful pop band in history and visionary doctors to help his idea reach fruition.
Few bands can rightly claim to have had an impact on the history of music, and even less can claim to have had a similar impact on the history of medicine. But The Beatles were a band that constantly pushed the boundaries of innovation in popular music, so it is no surprise to discover that their success helped fuel other advances – in this case the development of one of the most important diagnostic tools in the field of medicine.
Around the time the Beatles were finishing what many consider their most innovative album, Sergeant Pepper’s Lonely Heart’s Club Band, an electrical engineer working at Electric and Musical Industries (EMI) called Godfrey Hounsfield was on a weekend ramble in the English countryside, where he conceived a wondrous but ambitious idea. In fact Hounsfield’s idea for viewing and examining organs from outside the body was so ambitious it would require considerable financial investment to get it off the ground.
Thanks to The Beatles, however, whose record sales had almost doubled EMI’s profits since they had signed to its Parlophone label five years earlier, EMI had begun to invest a sizeable amount of money into funding bold research ideas. Within the space of five years Hounsfield’s idea would come to fruition, and few medical achievements would be received with such unreserved enthusiasm as would his invention of computed tomography (CT).
Hounsfield joined EMI in 1951, where he initially worked on radar and guided weapons. Among his achievements, his development of Britain’s first all-transistor computer, EMIDEC 1100, in 1958, stood out. EMI moved Hounsfield to its Central Research Laboratories after selling off its computer division in 1962, and assigned him to the project of designing a one-million-word immediate-access thin-film computer memory. This project failed for commercial reasons, but trusting Hounsfield’s creative capabilities, his supervisors gave him free rein to choose his next research project.
It was at this point that Hounsfield had his flash of inspiration on his countryside walk. Hounsfield was thinking about his radar research, in particular the problems of pattern recognition. Radar systems scan their surroundings by sending out radio waves from a central point and detecting patterns in the periphery. Why not try the reverse process, Hounsfield thought while walking, and study the central or interior pattern of an object from outside? Why not send beams through a parcel to find out what’s hidden inside?
“I thought, wouldn’t it be nice if I had many readings taken from all angles through a box,” said Hounsfield. “Wouldn’t it be nice if I could reconstruct in 3-D what was actually in the box from these random direction readings taken through the box?” The trick, thought Hounsfield, was to view the three-dimensional object as a series of cross-sectional scans, or slices, and he started working out mathematically how he could do this.
Hounsfield thought that X-rays could fulfil this purpose. X-rays are a phenomenally powerful tool for diagnosing fractures of bones. However, X-rays are less useful in diagnosing conditions affecting soft tissues like the brain, as the rays can’t distinguish one type of tissue from another – so they appear in X-ray photographs as a grey fog-like mass.
Hounsfield’s idea centred on the fact that the intensity of an X-ray beam reduces when it passes through an object – a process called attenuation. Different parts of the human body – for instance, bones and soft tissues like the brain – dampen X-rays differently. If you could observe the different attenuation patterns from a human object by directing an X-ray beam through it from different angles, hypothesized Hounsfield, you could distinguish between different types of tissues and reconstruct an image of a ‘slice’ of that object.
Hounsfield sent his research proposal to EMI under the title “An improved form of X-radiography”, in which he proposed that a series of X-ray exposures taken from different angles around an area of the body could construct a cross-sectional image of a ‘slice’ of that area. The different X-ray exposures could be detected by a sensing device that was always pointing towards the source of the gamma rays, and these readings would be digitized and fed to a computer to build up a crude picture of the material within the ‘slice’.
Hounsfield was unaware that the Austrian mathematician Johann Radon and the South African physicist Allan Cormack had already shown in theory that such an image could be obtained. For instance, Cormack devised a mathematical solution to measuring the tissue-density distribution within the body. On the basis of this, he proposed that X-rays could be taken from different angles around the brain or body, and accounting for the different effects of soft and dense tissues on X-rays, a computer could assemble these images into three-dimensional representations, but he did not take this idea further in terms of creating an instrument that could carry this out.
However, thanks to EMI’s deep pockets Hounsfield could begin to turn his idea into a working product. The project created enough excitement that the British Department of Health got involved through its Radiological Advisor Evan Lennon.
Hounsfield’s first experimental system used gamma rays from the radioactive element Americium to scan bottles or perspex jars filled with water and pieces of metal and plastic, and was “very much improvised”, as he recalled in his Nobel Lecture. A lathe bed provided the means for moving and rotating the gamma-ray source, and sensitive detectors were placed on either side of the bottles or jars. The scanning process took nine days and created 28,000 measurements, which took a high-speed computer two and a half hours to calculate and process. The images the computer created, though, were good enough to convince both EMI and the Department of Health to invest £6,000 each in the acquisition of an X-ray tube and a generator, which would reduce scanning time to nine hours. Hounsfield travelled across London by underground bringing bullock’s and pig’s brains fresh from the abattoir to his laboratory, and he produced the first pictures in which white and grey matter could be clearly differentiated.
The images from animal brains showed that the method worked, but Hounsfield needed to collaborate with a clinician to show whether it would work on human brains. The Department of Health’s Lennon tried in vein to find such a contact, but he came up against a wall of scepticism. “Why should I meet such a crank?” responded the first radiologist Lennon approached, and several more refused a request to meet Hounsfield.
Lennon persisted and eventually struck gold with a consultant radiologist called Jamie Ambrose, who was based at the radiology department of Atkinson Morley’s Hospital in Wimbledon, London. Ambrose was exploring ways of imaging the living brain using methods such as ultrasound and echo encephalography, and he agreed to meet Hounsfield. The first meeting did not go well. Ambrose found Hounsfield a difficult character; a man who was not very talkative, wary of explaining any details of his invention, and who responded to all the newest neurological images that Ambrose showed him with a dismissive “I can do better than that.”
As they made their goodbyes after what seemed an unfruitful meeting Ambrose handed Hounsfield a jar containing a brain with a tumour and asked him for some proof of his invention. When Hounsfield returned he brought a picture of the brain with him, which showed the tumour and even areas of bleeding within the tumour. Ambrose was instantly stunned.
Ambrose’s foresight and enthusiasm would prove to be a much-needed fillip for Hounsfield and his invention. Radiologists by and large were looking for ways in which to improve the resolution of X-ray images and reduce the time taken to get them. A new and revolutionary technique promising images of the brain, but that had less resolution and that took longer to acquire than existing methodologies wasn’t top of the wish list for most radiologists – or so they thought.
The prototype of what was called the EMI brain scanner was installed at Atkinson’s Morley Hospital and the first human patient was examined on October 1, 1971. (The EMI brain scanner was later renamed computed tomography, and the method is also known as computerized axial tomography, or CAT.) For the first CT scan Ambrose chose a woman in her early forties with a suspected brain tumour.
“There was a beautiful picture of a circular cyst right in the middle of the frontal lobe,” Hounsfield recalled, “and, of course, it excited everyone in the hospital who knew about this project.” After seeing the image Hounsfield and Ambrose felt, as the latter recalled, like footballers who had just scored the winning goal. The skull was no longer inaccessible from outside, the intricacies of the brain were now visible and they could distinguish between healthy and diseased tissue. Over the next few weeks they confirmed the capabilities of the scanner with around ten other patients, in whom they diagnosed and localized their brain diseases and thus made them accessible for surgical intervention.
When Ambrose presented these first clinical images at the Annual Congress of the British Institute of Radiology – the oldest radiological society in the world – on 20 April 1972 the audience was stunned. Seeing images of the brain that clearly showed lesions, tumours and haemorrhage instantly blew away all the scepticism radiologists had about the technique. These powerful images convinced radiologists that they were witnessing a new era in the detection and evaluation of disease.
The Department of Health purchased the first three EMI scanners produced and placed them in the Manchester Royal Infirmary, in Glasgow and at the Institute of Neurology in London. Two more scanners were sent to the Mayo Clinic and the Massachusetts General Hospital in the United States. In October 1972, Jamie Ambrose displayed and presented images from an EMI scanner to an audience of 2,000 medical doctors at the Chicago meeting of the Radiological Society of North America. He received a standing ovation, one of many that he and Hounsfield would receive over the next few years, as improvements in CT allowed sections of the body to be analysed, and the radiology literature began to reveal the full impact of the technique on the diagnosis and treatment of patients.
The invention of computed tomography provides an almost textbook example of how the progress of science relies as much on the belief and championing of ideas as it does on the quality of the idea itself. As one historian in the field says the risky realization of computer tomography could have happened nowhere else than in the United Kingdom at this particular time: “It is hard to imagine how the instrument would have gone into production without the support of a company like EMI. The combination of the Beatles’ success with the British system of research subsidies and the genius of one engineer broke the cash barrier and changed the face of modern medicine.”
Dixon, Adrian K.: Computed Tomography. Encyclopaedia of Life Sciences, 2001.
Hounsfield, Godfrey N.: Computed Medical Imaging. Nobel Lecture, December 8, 1979.
Husband, J. and Dombrowe, G. X-ray computed tomography – a truly remarkable medical development. British Journal of Radiology 78, 97–98 (2005).
Obituary Jamie Ambrose: The Guardian, May 10, 2006.
Obituaries Sir Godfrey Hounsfield: Independent, August 20, 2004; Telegraph, August 16, 2004; British Medical Journal 2004; 329; 687 (18 September); European Radiology 2004; 14; 2152–2153).
Petrik, Vladimir et al.: Godfrey Hounsfield and the dawn of computed tomography. Neurosurgery 58: 780–787 (2006).
Wells, P.N.T.: Sir Godfrey Newbold Hounsfield KT CBE. Biogr. Mems Fell. R. Soc. 51, 221–235 (2005).
See them all presented here.