Radiology is unique among medical specialties in that it is possible to trace its origins to a precise moment in history.

Professor Röntgen was a professor of physics at the University of Wurzburg in Germany when he discovered x-rays on November 8th 1895. Shortly thereafter he made a radiograph of his wife’s hand (Fig. 1). This finding and that of Becquerel, who was to discover radioactivity in early 1896, were forever to change the practice of medicine. Both discoveries were serendipitous. Röntgen was experimenting with a cathode-ray tube when he noticed fluorescence at a distance. This could only have been produced by much more penetrating radiation than cathode rays. Becquerel had put a piece of rock (pitchblende) on wrapped photographic film to examine the effect of sunlight on the mineral. He noticed that there was as much film blackening on an overcast day as in sunlight and deduced that the radiation came from the ore itself rather than being induced by the sun

In the aftermath of the first of these discoveries several investigators, both in Europe and North America, realized that they might have discovered x-rays before Röntgen. Inexplicably fogged photographic film, sometimes with "shadows" of overlying materials visible, betrayed the existence of penetrating rays. However, the connection with a then unknown form of radiant energy was not made by the individuals concerned (1,3-5).

As Pascal observed "chance favors the prepared mind." Röntgen was a careful experimentalist and his first paper, Eine neue Art von Strahlen, describes much of what we know about x-rays even today, including their use to explore human anatomy (Fig. 1) (8,9). Röntgen refused to gain financially from his discovery for which he was awarded the first Nobel Prize in physics in 1901 (10). Röntgen's paper, in the custom of the time, was circulated to colleagues in Europe. The one sent to Vienna was, by chance, discussed at a dinner party in the presence of a younger physicist whose father edited the daily newspaper Die Presse. Thus the news rapidly spread around the world. Reaction varied: Thomas et al. note that "The Lancet," the first English-language medical journal to report the discovery, "was initially skeptical, then factual and, by the end of January [1896], enthusiastic" (4,11).

Many scientists, professional and amateur, had apparatus in their laboratories similar to that used by Röntgen. They were immediately able to repeat his observations. Equally quickly many of these scientists made their equipment available to clinicians. In Great Britain Alan Campbell Swinton, an electrical engineer, "showed the ghastly pictures of his hands to his friends" on the evening of 7 January 1896 (11) and published a hand radiograph in the British Medical Journal on 25 January 1896 (4,11).

At Dartmouth College, New Hampshire, Edwin Frost, a professor of astronomy, was asked by Gilman Frost, his physician-brother, to radiograph a patient. The resulting image of the left forearm, exposure time 20 minutes, made on 3 February 1896, was the first radiograph made in North America (1,3,5). In Montreal a young man had been shot in the leg in a brawl on Christmas Eve 1895. He remained in pain and several surgical explorations had failed to locate the bullet. His surgeon, hearing about the new ray, took him to John Cox, the Professor of Physics at McGill University, on February 7th 1896. Cox made a radiograph, the exposure taking 45 minutes (12,13). The bullet was found lodged between the tibia and fibula. It was then successfully removed. The film was subsequently used in court, probably the first use of radiography in jurisprudence (14). Thus within a couple of months of its discovery, radiography was being used in many places in the English-speaking world, although a translation of Röntgen's paper had only appeared in Nature on January 16th, 1896 (4). Few discoveries have been adopted into medicine so quickly, or now might be.

At the turn of the century there were no specialties in medicine, as we now know them. Much early radiography was done by physicists and even town photographers (4,6,15). Meanwhile at his summer home in Baddeck, Nova Scotia, Alexander Graham Bell made many experiments with x-rays in 1896, which are recorded in his journals. He experimented with the telephone transmission of x-ray signals and while this may not have been the birth of teleradiology it was arguably the conception. Also Bell was later to be the first person to suggest the use of radium to treat cancer (16,17). As an example of the response to x-rays throughout the medical world Sir William Osler, perhaps the most notable physician of his time, was prompted to use x-rays experimentally in 1896. He embedded some gallstones in a beefsteak and made radiographs of it to see if the gallstones were detectable. The result was negative and we now know that only a small proportion of gallstones are opaque to x-rays.

Bone, soft tissue and dense foreign bodies provided the only contrast between materials in early radiography. Soon an orally administered contrast agent (bismuth nitrate replaced a decade later by barium sulphate) was given to study the alimentary tract (17,18). After much experiment an intravenous contrast agent was developed and, in 1927, marketed (Uroselectan, Schering AG) for urinary tract radiography (18).

For the first seven decades of the development of x-ray technology, progress was determined by technical feasibility (21). Important advances included the development of the more reliable Coolidge tube (1913) permitting shorter exposures; grids by Bucky and moving grids by Potter; conventional tomography by Ziedes de Plantes (1921) among many others; and cerebral angiography (1927) by Dr. Egas Moniz; while Dr. Werner Forssmann was the first to place a catheter in the heart (his own) (3,5).

Hevesy's description of the tracer principle, stemming from his collaboration with Rutherford beginning in 1912, was also to be important in physiological imaging (22).

Nor should it be forgotten that there was a price paid for this inventiveness, perhaps most strikingly revealed in the memorial to the x-ray martyrs in Hamburg (1,3-5,14).

Röntgen's work immediately had ramifications beyond the obvious clinical ones (14,23). Frau Röntgen saw in the radiograph made of her hand by her husband (Fig. 1) an intimation of mortality. The scientific interest in x-rays was, perhaps, exceeded only by the fascination they held for the public. To understand this interest we must leave the prejudices of our times and reflect on the society into which x-rays were announced. Radio had just been invented by Marconi and, with other modern inventions, suggested a boundless future of technological promise.

Queen Victoria was in the late years of her reign (1860-1907). Tannahill has noted that the peculiar mixture of sexual repression, perversion and the commercialization of sex that we associate with the Victorian era was in fact a worldwide phenomenon (24). Scientifically the world seemed a tidy place in the early 1890's. Atoms were perceived as tiny and indestructible spheres and the insights of modern physics were only to begin with the discovery of x-rays. A professor of physics of that era is even reputed to have told his students not to become physicists since “physics was over” (11).

Public perceptions surrounding the discovery of x-rays can be categorized as follows:
a) an immediate insight by the public into the potential for x-rays to change the practice of medicine;
b) an acceptance of x-rays as another aspect of technological mastery which was expected to relieve the human condition;
c) a prurient interest in the fact that these new rays might be used to see through materials, particularly clothing.

To Victorian sensibilities the ability of x-rays to penetrate women’s clothing in particular seems to have been very important. That society, certainly among its upper classes, set great store on propriety. X-rays as a means of invading privacy were the subject of many cartoons (25) and much comment in the popular press (14,23,26). The magazine Photography published some doggerel by Wilhelma shortly after the discovery that read in part:

Thro' cloak and gown - and even stays,
Those naughty, naughty Roentgen rays.

According to Harvey Graham, “A London firm rose to the occasion, and made a small fortune from the sale of x-ray proof underwear,” and in New York there was an attempt to legislate against the use of x-rays in opera glasses (27).

X-rays are invisible. Nevertheless, their discovery seems to have provided as penetrating a view of society at the time of their discovery as their clinical use offers a penetrating view of the skeleton.

The history of radiology is rooted in Röntgen's discovery. However, subsequently most if not all of the known physical energies have been explored for potential use in radiological diagnosis and treatment (Table 1). Radiological diagnosis is now often referred to as “imaging” reflecting the fact that it is no longer dependent only on x-rays. Nevertheless, the evolution of the radiological sciences has led to more profound changes than a simple multiplication of tools and the development of “imaging.”.

Table 1. Penetrating radiation and imaging and treating disease.

Early radiology was rooted in morphology, chiefly skeletal morphology. Beginning with nuclear medicine, powerful methods have been developed to examine bodily function. With this transformation has come a capacity not only to display the epiphenomena of disease (tumour masses, sinus tracks, etc.). Imaging tools have come to reveal bodily function including mechanisms of disease and the biology of treatments. This trend will almost inevitably increase.

Also, from a passive role in supporting medical diagnosis, radiology has returned to the bedside with interventional techniques that allow image-guided biopsy, drainages and treatments such as catheter-based cerebral aneurysm ablation (Fig. 3) as well as radionuclide therapy. Thus while the creation or use of images of internal structures plays a part in radiological practice, images are neither a necessary nor sufficient part of radiology to define its practice. Since this sub-specialty of interventional radiology is patient-friendly compared with traditional surgical methods (often requiring very short hospital admissions or none at all) it may also be important in addressing some of the political imperatives in the provision of treatments within the limited resources available.

Figs. 3: Selected views from a right internal carotid rotational angiogram showing (a) first coil passing from a micro catheter into a posterior communicating artery (PCA) aneurysm; ~) and (c) subtracted right internal carotid injection six-months later revealing stability of the aneurysm closure (images courtesy of Dr. Thomas Marotta, University of British Columbia (UBC) and Vancouver Hospital and Health Sciences Centre (\7HHSC) and reprinted from Annals, Royal College of Physicians and Surgeons of Canada 1995; 28: 470- 476 with permission).

 

There has been, in the last three decades, a rapid growth in the capacity of physicians to "image" disease: This has led to a bewildering number of acronyms describing new techniques (CT-computed tomography PET-positron emission tomography, SPECT-single photon emission computed tomography, MRI-magnetic resonance imaging, etc.).

This phase has been described as the medical-scientific phase of radiology, which began, with Hounsfield and Cormack's invention of CT (19). Thus also began the transition from analogue to digital imaging. However, unifying features of these techniques are:

a) Initially radiological images were two-dimensional representations of three dimensional tissues or organs. Newer imaging methods are sectional and reveal internal structure slice by slice in the way in which the slices of a loaf of bread, quite apart from their convenience also reveal the internal structure of a the loaf.
b) The use of high capacity microcomputers to reconstruct sectional or other images from complex data sets. The massive computational tasks involved in image reconstruction in PET, CT, SPECT and MRI would not be possible using a slide rule, certainly on a time scale relevant to human disease.
c) Another group of technologies also use computer image processing but do not yield sectional data (digital subtraction angiography (Fig. 3) and magnetic resonance angiography, etc.) It is important; nevertheless, to note that chest radiographs remain the most common radiological procedure.

Sound energy, used in wartime sonar, was soon applied to the study of disease. Clinical ultrasonography has become a highly developed tool for clinical diagnosis in Canada, probably because it is the cheapest of these capital-intensive technologies. Blume has written an interesting account about the sociology of their introduction (24).

This technology uses x-rays but in a new way. In present (third and fourth generation) machines a fan beam of x-radiation sweeps through 3600 while on the opposite side of the patient detectors provide a digital read out of the amount of radiation and hence the degree to which it has been attenuated. From the linear attenuation in multiple projections, it is possible to reconstruct a sectional display of body structure in terms of electron density.

The technology of nuclear magnetic resonance analysis has a long pedigree in physics, chemistry and biochemistry laboratories. Its use in life to obtain images of organ structure, variously weighted with "biochemical" data, stems from the realization that the use of gradient fields in intersecting coordinates will provide information, which encodes information about both the molecular environment of atoms and their spatial location. Thus images (to date of protons [H-1] and Na-23) have been created by algorithms similar to those used in CT. Some see the future of MRI as centering on proton imaging (already shown to be superior to CT in many contexts) while others are optimistic about the research and clinical applications of imaging other nuclides (C-13, F-19, Na-23) or of obtaining spectra revealing, for example, concentrations of phosphorus (P-31) metabolites in particular regions of organs such as the brain (Fig. 6). MRI techniques can be enhanced by intravenous injection of agents analogous to the contrast (iodine containing) materials used in radiography.

Fig. 6: Functional data (in yellow) in orthogonal planes from a single shot gradient-echo echo-planar internal speech experiment superimposed on Ti- weighted MRI images. The yellow pixels reflect statistically significant (p<O.05) signal increases in the inferior frontal region of the dominant hemisphere in this left-handed patient, who had been asked to think (but not speak aloud) of as many words as possible starting with a given letter of the alphabet (image courtesy of Dr. Bruce Forster, UBC and VHHSC).

The unique characteristics of annihilation radiation make it possible to carry out quantitative sectional imaging of the distribution of radionuclides which decay by ß emission. These radionuclides emit positrons (ß+ particles or positively charged electrons) that collide with an electron resulting in a mutual anibilation and the transformation of their rest mass into two gamma rays emitted at 180 degrees to each other. The detection of these rays permits reconstruction of the tissue distribution of the parent radionuclide slice by slice through an organ or in the body (Fig. 4).

The fact that the only externally detectable isotopes of some biologically important nuclides, namely carbon (C-11), nitrogen (N-13) and oxygen (0-15) decay by ß+ emission reflects the potential importance of PET. These atoms make up nearly all of bodily tissues except bone. As a method it has been described as providing "autoradiographs in life."

Fig. 4: Transverse PET images of the brain: (a) F-18 fluorodopa (FD) and C-Il raclopride scans in a patient with idiopathic Parkinson disease. The first reveals clear evidence of damage to pre-synaptic neurons while the second (sensitive to D2 post-synaptic receptor concentration) shows compensatory up-regulation, and (b) serial scans with FD in a Parkinson disease patient who had undergone fetal cell transplant The changes were matched by some clinical improvement.

This technique uses the single gamma rays emitted by other radionuclides used in nuclear medicine (Tc-99m, 1-123, In-ill etc.). One or more gamma camera heads is mounted on a rotating gantry and circles the patient. Using computed algorithms like those used in CT or PET, an image of tracer distribution in multiple organ sections is obtained (Fig. 5). Ultrasonography (US): From its beginnings in the sonar detection of submarines, US has undergone increasing technological sophistication and now arrayed transducers provide images in real time in which the signal strength is proportional to reflection from tissue interfaces. Moreover, the use of Doppler shift analysis permits recognition of the velocity and direction of blood flow. US are widely used in pregnancy for both mother and fetus since no exposure to ionizing radiation is involved.

Fig. 5: (a) A transverse SPECT image of a right apical lung mass, which demonstrates hyper metabolism of F-18 fluorodeoxyglucose. (b) A transverse CT scan at the same site revealing the 3 cm. diameter mass. (c) Composite anatomical and functional images of the mass, which subsequently proved to be a large cell lung carcinoma. (images courtesy of Dr. D. Worseley, UBC and VHHSC).

The naturally occurring magnetic fields associated with electrical activity in the brain and heart can be displayed using multiple magnetometers in a MEG device.

This technique has been found to be valuable in those patients with surgically treatable epilepsy (particularly partial complex seizures) and cardiac arrhythmias as well as in planning brain surgery to avoid damaging critical areas of the cerebral cortex.

Techniques have now been developed for digital rather than photographic recording of conventional x-ray images (28). This technology will be relatively easily introduced since the investment may be offset in part by savings in film purchase and storage but more importantly by efficiencies in film retrieval by other clinicians. Digital images will provide more latitude in exposure and some potential for image processing, e.g. edge enhancement.

Hospitals have in general been slow to adopt modern information technology. This is particularly damaging in radiology departments because they generate about 80% of the data bits produced in a typical hospital. This is now changing. As fluid systems for image management and transfer result, there will be fewer radiographs lost and shorter waits at film handling and transmission of images both locally and by telemetry, while computer assisted film interpretation and decision support are already on the horizon and will be readily introduced into the innovative, computer dependant world of modern radiology.

In reflecting on these technologies it is important to distinguish between their use as tools to study mechanisms of disease and their use in a hospital or clinic to diagnose and follow the time course of an individual patients illness. It is sometimes argued that certain radiological methods might serve only for research. More likely all of these tools will have a clinical role but their degree of dissemination outside major and academic institutions will vary for each.

To the frustration of hospital administrators and political paymasters each imaging method has strengths and weaknesses. No single method will solve all diagnostic problems. Many radiologists, however, believe that MRI (for reasons of diagnostic power) and ultrasonography (for reasons of simplicity, relative cheapness, effectiveness, and potential use in less-developed nations) along with image-guided interventions will be the defining radiological technologies of the foreseeable uture. Plain film radiography particularly of the chest will, however, be a large part of the work in a radiological or imaging service (in both hospital or ambulatory care facilities for the foreseeable future).

The resolution achievable by the different imaging methods may be classified as spatial, contrast or temporal. Spatial resolution is the ability of a system to resolve anatomic detail. Contrast resolution is the ability of a system to be used to distinguish one tissue from another or diseased from normal tissue. Temporal resolution is the ability of a system to reflect either changing physiological events such as cardiac motion or disease remission or progression as a function of time. High spatial and temporal resolution is particularly achievable by radiography, CT, MRI and US. High contrast resolution is particularly achievable using PET, SPECT, MRI and US. MRI is probably the most powerful single tool in all three contexts. However, there is no prospect of it supplanting PET for example, in displaying cell-surface receptors in the human brain using such probes as F-18 fluoro-dopamine (Fig. 4).

The manufacturers of radiological machines and contrast agents clearly have a close relationship with the practice of radiology. The history and sociology of the introduction of radiological technologies has been examined by Blume (29). His thesis that radiologists and industry have conspired in deciding which technologies survive is a point of view. However it must be set against biological and physical limitations, which determine what, is and is not possible in imaging and treating disease (Table 1).

Radiological methods are an obvious resource for the teaching of anatomy, physiology and gross pathology. This resource tends to have been under-used in the past but some university departments of anatomy are now coming to depend very heavily on radiological teachers and not just in the matter of sectional structure. Medical schools, however, have often been slow to respond to the rapid evolution of radiology even though Barjon wrote before 1918 that " if the radiologist ought to be a physician, it would be well also for the physician to be, in a lesser degree also a radiologist." (30).

There is not an inexhaustible range of physical energies available with which to image the body or treat disease. Indeed the next two decades will see less radical change and instead the wider application and better understanding of the roles of the technologies described. However, other developments taking place or being investigated for use in radiology include:

The idea of using light to image the interior of the body as distinct from its surface is not new. Intense light has been used to trans-illuminate breast tissue and the extremities in a darkened room. Certainly breast cancers have been found to be visible against normal breast tissue. However, to transmute this method into a recordable examination of value in diagnosis poses some problems in signal analysis.

This method, which also requires high field magnets, is less eloquent than NMR in the laboratory. Some doubt that the technique is possible in life but research in this context is underway and if the history of radiology teaches one anything it is that what is preposterous today is often in clinical use tomorrow.

Electrical impedance imaging: It has been known for over a decade that weak electrical currents are absorbed differently by different tissues. The technique has been slow to find clinical applications despite its relative simplicity and low cost. However, sectional impedance imaging has now been proposed and a commercial device produced capable of detecting small breast cancers with a high degree of sensitivity and specificity. The technology may have a role as a precursor to mammography.

While the evolution of radiologic technology will continue, issues of health care policy are becoming central to the future of the radiological sciences. The assessment of new technologies must be rigorous while avoiding technological nihilism. Also there is an over-riding need to eliminate unnecessary and redundant examinations (appropriateness). Nevertheless radiology has a potential to contribute to solving some of the issues in patient care and education, which face medicine in the future, particularly in information management and in providing low cost and patient-friendly diagnostic and treatment procedures. Indeed developing image-guided surgical methods are likely to contribute positively to our ability to provide care using a realistic fraction of societies wealth. To that degree, funding radiological methods in the future may come to be seen less as a resource-intensive burden and more as a strategic solution to delivering cost-effective care.

1. Grigg ERW. The trail of the invisible light. Illinois: Thomas, 1965.
2. Donizetti P. Shadow and substance: the story of medical radiography. Oxford:
Pergammon Press, 1967
3. Eisenberg RL. Radiology: an illustrated history. St. Louis: Mosby, 1992.
4. Thomas AMK, 'Sherwood I, Wells PNT. The invisible light: 100 years of
medical radiology. Oxford: Blackwell Science Ltd., 1995.
5. Gagliardi RA, McClennan BL. A history of the radiological sciences: diagnosis.
Reston, VA: Radiology Centennial Inc, 1996.
6. Aldrich I. Röntgen and the discovery of x-rays. In: Aldrich JE, Lentle BC,
editors. A new kind of ray: The radiological sciences in Canadal89S-1995. -
Montreal: The Canadian Association of Radiologists, 1995, p 5.
7. Dutreix I, Dutreix A. Henri Becquerel (1852 - 1908). Med Phys 1995; 22: 1869-
1875.
8. Bruwer AJ. Classic descriptions in diagnostic radiology. Vol.1. Illinois:
Thomas, 1964.
9. Klickstein F's. On a new kind of rays: a bibliographic study. St. Louis:
Mallinckrodt, Classics of Radiology, Vol.1, 1966.
10. Knutsson F. Röntgen and the Nobel Prize. Acta Radiologica Diagnosis 1969; 8:
449A60.
11. Posner E. Reception of Röntgen's discovery in Britain and the U.S.A. Brit Med
J1970; 4; 357-360.
12. Cox J, Kirkpatrick RC. The new photography with report of a case in which a
bullet was photographed in the leg. Montreal Medical Journal 1896; 24: 661.
13. Cohen M. Canada's first clinical x-ray. In: Aldrich JE, Lenfle BC, editors. A
new kind of ray: The radiological sciences in Canada 1895-1995. Montreal: The
Canadian Association of Radiologists1 1995, p 17.
14. Kevles BH. Naked to the bone: medical imaging in the twentieth century.
New Brunswick, New Jersey: Rutgers University Press, 1997.
15. Aldrich J. X-rays in Nova Scotia. In: Aldrich JE, Lentle BC, editors. A new
kind of ray: The radiological sciences in Canada 1895-1995. Montreal: The
Canadian Association of Radiologists, 1995, p 77.
16. Bell AG. Radium and cancer. Science 1903; 18: 1555.
17. Aldrich J. Alexander Graham Bell. In: Aldrich JE, Lentle BC, editors. A new
kind of ray: The radiological sciences in Canada 1895-1995. Montreal: The
Canadian Association of Radiologists, 1995, p 21
18. Williams FH. The x-ray in medicine. Med Rec NY 1896; 49: 665.
19. Cannon WB. The movements of the stomach studied by means of the
Roentgen ray. Amer J Physiol 1898; 1: 359-
20. Urich K. Successes and failures in the development of contrast media. Berlin:
Blackwell Wissenschafts-Verlag, 1995.
21. Schwiertt: G, Kirchgeorg M. The continuous evolution of medical x-ray
imaging. Electromedica 1995; 63: 2-8 and 3440.
22. Levi H. George Hevesy and his concept of radioactive indicators in
retrospect. Eur J Nuci Med 1976; 1: 3-10.
23. Lentle BC. X-rays as metaphor. In: Aldrich JE, Lentle BC, editors. A new kind
of ray: The radiological sciences in Canada 1895-1995. Montreal: The Canadian
Association of Radiologists, 1995, p 287.
24. Tannahill R. Sex in history. New York: Stein and Day, 1980, p 348.
25. Globe, Toronto, February 27, 1896.
26. Connor JTH. The adoption and effects of x-rays in Ontario. In Aldrich JE,
Lentle BC, editors. A new kind of ray: The radiological sciences in Canada 1895-
1995. Montreal: The Canadian Association of Radiologists, 1995 (reprinted), p
119.
27. (Quoted in) Jack D. Rogues, rebels and geniuses. Toronto: Doubleday, 1981, p
310.
28. Lee DL, Cheung LK, Jeromin Ls. A new digital detector for projection
radiography. SPIE, Physics of Medical Imaging 1995; 2432: 237-241.
29. Blume 55. Insight and industry: On the dynamics of technological change in
medicine. Cambridge1 Mass: M~ Press, 1992.
30. Barjon F (trans. Honeij IA). Radio-diagnosis of pleuro-pulmonary affections.
New Haven: Yale University Press, 1918.
31. Kotre CJ. Electrical impedance tomography. Brit J Radiol 1997; 70: S200 – S205.
32. www.TSCAN.org
33. Hebden JC, Delpy DT. Diagnostic imaging with light. Brit J Radiol 1997; 70: S206 – S214.