Born from Physics: How Nuclear Medicine Began (1896–1958)

On 12 November 1936 the president of MIT, the physicist Karl Compton, gave a lunchtime talk at Harvard Medical School called “What Physics Can Do for Biology and Medicine”. In the audience was Saul Hertz, the 31-year-old head of the thyroid clinic at Massachusetts General Hospital. When the talk ended he asked one question: “Could iodine be made radioactive artificially?” [1]

Compton thought it could. Less than five years later, on 31 March 1941, Hertz gave a drink of radioactive iodine to a woman with an overactive thyroid. It was the first time a radioactive drug was aimed at a single organ, relying on that organ’s own appetite for the element [1]. Every radioligand therapy today, from Lutathera to Pluvicto, follows the same logic.

Hertz’s question could not have been answered ten years earlier. It needed three ideas, all from physics, that took forty years to fall into place. This article tells how they did, and how nuclear medicine then learned to see.

Idea one: radiation exists (1896–1899)

In February 1896 the French physicist Henri Becquerel was testing a hunch. X-rays had been discovered a few weeks earlier, and he suspected that glowing (phosphorescent) uranium salts might absorb sunlight and give it back out as X-rays. He wrapped photographic plates in black paper, put uranium crystals on top and left them in the sun. The plates darkened, and on 24 February he reported the result [2].

Then the Paris weather turned grey. Thinking he needed bright sun, Becquerel put the crystals and plates away in a drawer. On 1 March he developed the plates anyway, expecting a faint image. It was strikingly clear. The uranium was emitting penetrating radiation with no sunlight at all, and in May he showed that the element uranium itself was the source [2].

Marie and Pierre Curie took up the “uranium rays” and in 1898 found two new, far more intensely radioactive elements, polonium and radium. Marie Curie coined the word radioactivity. In 1903 Becquerel and the Curies shared the Nobel Prize in Physics [2]. Their names live on in the units of activity: today’s becquerel, and the older curie.

In 1899 Ernest Rutherford showed that uranium’s radiation came in at least two kinds, which he named alpha and beta. Alpha rays were stopped by a thin sheet of material; beta rays penetrated much further [3]. More than a century later these are still the two main kinds of particle used to treat cancer. Beta emitters such as lutetium-177 (¹⁷⁷Lu) power today’s radioligand therapies; alpha emitters such as actinium-225 (²²⁵Ac) are coming next. Article 5 of this series explains why the difference matters so much.

Idea two: radioactive atoms can be traced (1913)

The second idea came from a failure, told in article 1. George de Hevesy could not separate “radium D” from lead because it was lead, a radioactive isotope of it. By 1913, working with Fritz Paneth, he had turned that into a method: add a little radioactive lead, and its radiation reveals where all the lead goes. A decade later he followed lead through a living bean plant [4]. The work won him the Nobel Prize in Chemistry for 1943, awarded in 1944 [5].

The tracer principle had one big limitation. Nature’s radioactive atoms are almost all heavy elements, such as uranium, radium, polonium and radioactive forms of lead and bismuth. The body barely uses these elements. To trace or treat real biology, scientists needed radioactive versions of the elements life is built from: iodine, phosphorus, sodium, carbon. Nature did not supply them.

Idea three: radioactive atoms can be made (1930–1938)

The answer came from Berkeley, California. Ernest O. Lawrence had arrived at the University of California in 1928. In the autumn of 1930 he built the first working model of a new kind of particle accelerator, the cyclotron, which whirled charged particles round a spiral path to ever-higher energies. It was about five inches (13 cm) across, and on 2 January 1931 it accelerated protons to 80,000 electronvolts. That August he founded the Radiation Laboratory, and in 1939 he received the Nobel Prize in Physics [6].

Early in 1934, in Paris, Irène Joliot-Curie (Marie Curie’s daughter) and Frédéric Joliot bombarded aluminium with alpha particles. The aluminium stayed radioactive after the alpha source was taken away. They had made a radioactive isotope of phosphorus, the first artificial radioactivity [7]. It earned them the 1935 Nobel Prize in Chemistry [8]. Together with the cyclotron, it meant that, in principle, a radioactive form of almost any element could now be made.

Medicine moved quickly. Lawrence’s brother John, a physician, joined him in Berkeley in 1935. He first injected cyclotron-made phosphorus-32 (³²P) into mice with leukaemia, and in 1937 he used it to treat a patient with polycythaemia vera, a disease in which the bone marrow makes too many blood cells. The Berkeley laboratory describes this as the first successful treatment of a human disease with radioisotopes [6]. Phosphorus is used by every cell, though, so this was a broad treatment rather than a targeted one. The first truly targeted therapy came four years later.

The cyclotron also produced a discovery nobody recognised as medical at the time. In February 1937 Lawrence posted a piece of molybdenum foil, part of his cyclotron’s deflector, to the physicist Emilio Segrè in Palermo. Segrè and the chemist Carlo Perrier found in it element 43, the first chemical element ever made artificially [9]. It was later named technetium, from the Greek for “artificial”. In 1938 Segrè and Glenn Seaborg found a form of it with a six-hour half-life [10]. That form, technetium-99m (⁹⁹ᵐTc), is today the most widely used radionuclide in medicine.

Hertz’s question: radioiodine, 1936–1946

Hertz’s 1936 question had an obvious medical logic. The thyroid gland takes up iodine from the blood to make its hormones, and an overactive thyroid takes up even more. A radioactive iodine would therefore travel straight to the gland.

Hertz teamed up with Arthur Roberts, a young MIT physicist. Their first radioiodine was iodine-128, made with a radium–beryllium neutron source at MIT. Its half-life was only about 25 minutes, too short for treatment but enough for experiments in rabbits, which showed that the thyroid took up the iodine as expected. They published in 1938 [1].

For treatment they needed longer-lived iodine, and that needed a cyclotron. Compton and colleagues won a $30,000 grant from the Markle Foundation to build one at MIT, and it began working in November 1940. It produced a mixture of about 90% iodine-130 (half-life 12.4 hours) and 10% iodine-131 (about 8 days) [1].

On 31 March 1941 Hertz’s first patient drank 2.1 millicuries (about 78 MBq) of that mixture. Seventeen days later she received a second dose of 1.3 mCi (about 48 MBq), and her metabolic rate fell to normal. Between 1941 and 1943 Hertz and Roberts treated 29 patients, of whom 20 were considered cured. In 1943 Hertz volunteered for military service, and a colleague, Earle Chapman, took over the clinic and continued the treatments. After the war both groups’ results appeared back to back in the same 1946 issue of JAMA [1].

Why call this the first theranostic? Because iodine-131 does two jobs at once. Every decay releases a beta particle (in about 89% of decays with a maximum energy of 606 keV), which deposits its energy within a few millimetres of thyroid tissue. About 81% of decays also give off a gamma ray of 364 keV, which escapes the body and can be counted or imaged from outside [11]. Targeting comes from the body’s own chemistry, beta particles do the treating, and gamma rays let doctors see where the drug went. That is the template that ¹⁷⁷Lu-based radioligand therapies follow today, with an engineered molecule in place of the thyroid’s appetite for iodine.

Isotopes for everyone: Oak Ridge, 1946

Cyclotrons make radioactive atoms a few at a time. Nuclear reactors, built in secret during the Second World War, make them by the bucketful: any material placed inside is bathed in neutrons, and many of its atoms are turned into radioactive isotopes.

After the war the United States turned that capacity outward. On 2 August 1946, at a small ceremony at Clinton Laboratories (later Oak Ridge National Laboratory) in Tennessee, scientists sent the first official shipment of a reactor-made radioisotope: carbon-14, bound for the Barnard Free Skin and Cancer Hospital in St Louis [13]. In its first year of production the Oak Ridge X-10 Graphite Reactor made more than 1,000 shipments of 60 different radioisotopes, chiefly iodine-131 and phosphorus-32 [14].

Radioiodine was no longer a rare product of a single university cyclotron; any hospital could order it. The same pattern holds today. Lutetium-177, the workhorse of modern radioligand therapy, is made in reactors, and isotope supply is a strategic question for every country that wants to offer these treatments. Article 6 covers how medical isotopes are made now.

Half-lives from LNHB decay data [11], except iodine-128 and iodine-130 [1].

Learning to see (1949–1958)

By the late 1940s doctors could detect radioiodine in the thyroid from outside the neck. Mapping the gland’s shape was another matter. With a detector positioned by hand, tracing the outline of a thyroid point by point took 60 to 90 minutes [18].

In 1949 Benedict Cassen, a physicist at the University of California, Los Angeles, automated the job. His rectilinear scanner moved a detector back and forth across the body, recording the radiation at each point and building an image line by line. Colleagues later said he got the idea for the printed readout from a stock ticker in his broker’s office. Clinical studies followed in 1950, and two of his colleagues began commercial production, first in a garage [18], [19]. The scanner was a big step forward, but it was still slow: each image was collected one spot at a time.

Hal Anger, an engineer at Berkeley’s Donner Laboratory, took a different approach: capture the whole field of view at once. In 1952 he built a pinhole camera using a sodium iodide crystal, which gives off a tiny flash of light when a gamma ray hits it, and a photographic plate. By late 1956 he had the design that bears his name: a crystal viewed by seven photomultiplier tubes, with electronics that work out where each flash occurred from how its light is shared between the tubes [20]. He published it in 1958 and demonstrated it at the Society of Nuclear Medicine’s annual meeting that June [20], [21]. The first commercial Anger camera was delivered in 1962. Within a few decades rectilinear scanners had all but disappeared. Gamma cameras, including the heads of modern SPECT scanners, still work on Anger’s principle [20].

The camera needed a radionuclide to match, and one arrived at almost the same moment. In the late 1950s Walter Tucker, Margaret Greene and Powell “Jim” Richards at Brookhaven National Laboratory, working on a different separation problem, built the technetium-99m generator [22]. It is a small shielded column holding molybdenum-99, which has a half-life of 66 hours. As the molybdenum decays it produces ⁹⁹ᵐTc, which can be washed out with salt water each morning [11], [22]. The long-lived parent can be shipped to hospitals, and the six-hour daughter is made fresh on site. Richards proposed technetium for medical imaging at a conference in Italy in 1960 [22].

Technetium-99m’s gamma ray, about 140 keV, suits a gamma camera well, and its six-hour half-life keeps the patient’s radiation dose low [11]. Today it is used in about 80% of nuclear medicine imaging procedures worldwide [22]. The Anger camera and the technetium generator, both developed in the 1950s, became the foundation of diagnostic nuclear medicine for the rest of the century.

What the founders left us

Almost every element of a modern radioligand therapy can be traced back to this period:

• Targeting comes from Hevesy’s tracer principle. A molecule decides where the radionuclide goes. In 1941 that molecule was simply iodide; today it is an engineered molecule built to bind a protein on cancer cells.
• Beta therapy with built-in imaging comes from Hertz. Iodine-131 treated with beta particles and could be followed by its gamma rays. Lutetium-177 does the same.
• Isotope supply comes from Oak Ridge. Therapy depends on reactors and accelerators that hospitals do not own, and that dependency remains one of the field’s main bottlenecks (article 20).
• Imaging infrastructure comes from Cassen, Anger and Brookhaven. The gamma camera and the technetium generator built the diagnostic side of the specialty that therapy now relies on.
• Alpha particles, which Rutherford named in 1899, are the field’s next frontier.

What comes next in this series

By 1958 nuclear medicine had its basic tools, but it was still mostly a thyroid specialty, and its images were flat. Article 3 follows the next six decades: the move to three dimensions with SPECT and PET, the first glucose scans with FDG, the detour of antibody-based radiotherapy, and the two lines of research, in neuroendocrine tumours and in prostate cancer, that led to Lutathera and Pluvicto.

Part III then takes up the physics this article touched on: what “half-life” and “becquerel” really mean (article 4), why alpha and beta particles damage cells so differently (article 5), and where medical isotopes come from today (article 6).

References

1. Fahey FH, Grant FD, Thrall JH. Saul Hertz, MD, and the birth of radionuclide therapy. EJNMMI Phys. 2017;4:15. doi:10.1186/s40658-017-0182-7
2. Tretkoff E. March 1, 1896: Henri Becquerel discovers radioactivity. APS News (This Month in Physics History), March 2008. aps.org
3. Rutherford E. Uranium radiation and the electrical conduction produced by it. Philos Mag. 1899;47:109–163. doi:10.1080/14786449908621245
4. Lindau Nobel Laureate Meetings. George de Hevesy — research profile. mediatheque.lindau-nobel.org/laureates/de-hevesy/research-profile
5. The Nobel Prize in Chemistry 1943: George de Hevesy. NobelPrize.org. nobelprize.org/prizes/chemistry/1943/summary
6. Lawrence Berkeley National Laboratory. Ernest Lawrence and the cyclotron; John Lawrence and the beginnings of nuclear medicine (75th anniversary exhibit). lbl.gov
7. Joliot F, Curie I. Artificial production of a new kind of radio-element. Nature. 1934;133:201–202. doi:10.1038/133201a0
8. The Nobel Prize in Chemistry 1935: Frédéric Joliot and Irène Joliot-Curie. NobelPrize.org. nobelprize.org/prizes/chemistry/1935/summary
9. Lindau Nobel Laureate Meetings. Emilio Segrè — research profile (the discovery of element 43). mediatheque.lindau-nobel.org/laureates/segr/research-profile
10. Segrè E, Seaborg GT. Nuclear isomerism in element 43. Phys Rev. 1938;54:772. doi:10.1103/PhysRev.54.772.2
11. Laboratoire National Henri Becquerel (LNHB/CEA). Table de Radionucléides: decay data for ¹³¹I, ³²P, ¹⁴C, ⁹⁹Mo and ⁹⁹ᵐTc.
12. Greenspan BS. Theranostics in perspective: white paper — a correction [letter]. J Nucl Med. 2025;66(11):1846. doi:10.2967/jnumed.125.271218
13. Oak Ridge National Laboratory. ORNL isotopes: 80 years, infinite possibilities. 12 August 2026. ornl.gov
14. US National Park Service, Manhattan Project National Historical Park. Oak Ridge X-10: 1946 — radioisotope production. nps.gov
15. Yang W, Kang F, Chen Y, et al. Landscape of nuclear medicine in China and its progress on theranostics. J Nucl Med. 2024;65(Suppl 1):29S–37S. doi:10.2967/jnumed.123.266968
16. China clear to decommission its first research reactor. World Nuclear News, 20 November 2019. world-nuclear-news.org
17. Nuclear milestones: China's nuclear industry shifts from military to civilian use. Beijing Review, 27 November 2018. bjreview.com
18. Nuclear medicine pioneer: Benedict Cassen — the rectilinear scanner and an enduring legacy of education and research. J Nucl Med. 1998;39(5):16N. jnm.snmjournals.org
19. Blahd WH. Ben Cassen and the development of the rectilinear scanner. Semin Nucl Med. 1996;26:165–170. doi:10.1016/S0001-2998(96)80021-3
20. Tapscott E. Nuclear medicine pioneer: Hal O. Anger, 1920–2005. J Nucl Med Technol. 2005;33:250–253. tech.snmjournals.org
21. Anger HO. Scintillation camera. Rev Sci Instrum. 1958;29:27–33. doi:10.1063/1.1715998
22. Brookhaven National Laboratory. Celebrating the 60th anniversary of technetium-99m. 24 October 2018. bnl.gov
23. Society of Nuclear Medicine and Molecular Imaging. SNMMI history. snmmi.org