Many of you in the audience have made substantial contributions to the current extensive use of radiation and radioisotopes in the biomedical sciences. Most of these contributions were not directly motivated by biomedical objectives. Nevertheless, your contributions to detector development have made possible the extensive and varied applications that radiations and radioisotopes find in the life sciences today.

The particle and photon detectors usually discussed at this symposium are predominantly those upon which Nuclear Physics depends, namely: photomultiplier tubes, image intensifier tubes, scintillators, semiconductor detectors and spark chambers. These are also the basic detectors employed in Radiobiology, Photobiology and Nuclear Medicine.

The fundamental detection mechanisms and many of the measurements problems in Physics and Biology are similar. There are also important differences. For the discussion today, I have chosen specific applications that show these similarities and differences. The examples also appear especially significant to me for their present and potential biomedical importance.

Most of the biomedical detectors are used to identify, localize or measure radioactive

materials which have been introduced into the body. These materials may have been administered deliberately so that function and metabolic studies can reveal the way particular molecules are handled biologically. In diagnosis, malfunctions can be demonstrated by anomalous distributions or movements of the introduced radioisotope. In treatment, radioisotopes are used to destroy cells either to reduce function or in an effort to eliminate tumor tissue

At very low levels, radioactivity is also present naturally in living systems, as it is in all materials and additional radioactivity in varying amounts comes in from environmental contamination and from accidental exposures. Identification and measurement of this sort of radioactivity may be extremely important.

SCANNING

The development of scintillation counting as a practical laboratory tool gave us a way to detect externally the presence and distribution of radioactivity in the human body. In 1950 Cassen! developed the first radioisotope scanner at the University of California, Los Angeles. Cassen's scanner was used to visualize thyroid gland lesions by mapping the uptake of radioiodine. His rectilinear scanning device was a collimated scintillation counter that moved back and forth across the thyroid region detecting and recording the presence of radioactivity as it passed over it. The first scanners were useful for the thyroid because it lies near the surface and absorbs radioiodine readily. Since then the introduction of pulse height analysis, large crystals, efficient multi-aperture collimator designs and a variety of developments in image display techniques have made scanning a reliable clinical procedure for other body regions.

Rectilinear scanners are still the most widely used type of scanner today. More recently, "stationary" scanners have been developed that operate as cameras. These cameras visualize the entire distribution of radioactivity in their field of view at a given moment. Unlike rectilinear scanners, the speed of image formation in cameras is not limited by mechanical

considerations. Because of this increased amounts of short half-life radionuclides can be used to improve image quality and at the same time reduce the dose to a patient. Their capability for taking pictures in rapid sequence makes possible dynamic studies of organ function where the isotope distribution is changing rapidly.

The scintillation camera was first developed by Anger2 at the Lawrence Radiation Laboratory in Berkeley. Since then the Autofluoroscope--an adaptation of Anger's early work--has been developed by Bender and Blau3 at Roswell Park in Buffalo. The Anger camera uses a thin sodium iodide crystal which is viewed by an array of photomultiplier tubes, as illustrated in Fig. 1. A scintillation in the crystal generates a set of photomultiplier tube output signals. The distribution of responses among the tubes depends on the location of the scintillation in the crystal and is unique for that location. A computing circuit transforms the set of phototube responses into beam deflection in an image-readout oscilloscope. At the same time, all the phototube signals for each scintillation are summed. Whenever the total signal falls within the photo peak of the gamma ray spectrum of interest, the oscilloscope beam is gated on and a point flash of light appears. The display is generally photographed by a time exposure of the oscilloscope screen.

These cameras require the performance that is characteristic of the better scintillation counters. In addition, the need for good position resolution introduces competing parameters that affect the overall system performance. For example, thick scintillators increase the number of primary photoelectric and multiple Comptonphotoelectric interactions, both of which are normally recorded as image dots, thereby enhancing detection efficiency. On the other hand, thick crystals tend to degrade the position resolution. This is because multiple interactions distribute light emissions over a relatively large region of the crystal. The computing circuit places the display flash for this interaction at the luminous center of the scintillations, which depending on the primary gamma energy, may be at some distance from the initial interaction point.

The low energy statistical variations and high energy collimation problems limit the most useful range of gamma cameras to about the range of 70 KeV to 400 KeV.

Bender and Blau employ in their Autofluoroscope a 2 in. thick mosaic of 3/8 in. diameter

Nal crystals, which increases sensitivity to medium and high energy gammas. The electronics of the original Autofluoroscope and Anger's scintillation camera are essentially the same. However, a new version of the Autofluoroscope4 utilizes a rank and file system of photomultipliers and light pipes. In this system a scintillation in one of the crystal elements generates simultaneous pulses from the two phototubes monitoring respectively the row and the column of the crystal elements involved. These pulses are then applied to the oscilloscope as positioning signals. The inherently digital nature of this method makes it very suitable for core memory storage.

diminution in the size of the left kidney, as demonstrated by the Neohydrin picture in Fig. 2C, and marked improvement in kidney function, as shown by the stop-motion hippuran excretion study.

Special forms of radioisotope cameras have been developed to image distributions of positron emitters. Positron cameras use coincidence detection of the two positron annihilation gamma rays, which makes conventional collimators unnecessary. Elimination of the collimator also improves the detection efficiency by increasing the usable crystal surface area. Long exposures of subjects containing very little activity is possible because the background in the absence of positron emitters is only a few counts per hour. Perhaps the most important property of positron cameras is their substantially uniform response to emitters at different depths in tissue, which allows them to visualize deep seated tumors with the same sensitivity as those on the surface.

The

Simplified gamma cameras employing image intensifier tubes have been developed for use with low-energy X-ray and Y-ray emitters. These systems comprise essentially only a collimator, a shielded image converter and means for magnifying and recording the image.5 Fig. 3 shows a system used by Ter-Eogossian, Washington University. collimators operate efficiently at low energies to reduce background due to photon penetration of aperture separations. However, the absence of pulse height analysis allows photons that are scattered into the collimator apertures to appear as background in the image.

Position resolution for low-energy emitters is in better shape statistically in the image converters than in the cameras because the thin phosphor and the photocathode are in direct contact. All the photoelectrons that generate a particular display dot come from the same small area of the photocathode and are focused together onto the output phosphor. The brightness of the spot varies statistically and with gamma energy, but its position does not.

Resolution suffers with currently available large image tubes because of face plate curvature which places most of the photocathode surface rather far from the collimator. However, specially constructed image tubes should enable this system to give good results for low-energy emitters.

A few spark chambers have been developed that replace the image intensifier tube in simple low-energy gamma cameras. In these cameras spark

discharges are initiated by interactions with collimated gamma rays and are imaged directly on photographic film.

MICROSCOPIC OBSERVATIONS

Studies which involve the use of minute tracer amounts of radioelements require radioactivity mapping that is beyond the resolution and sensitivity capabilities of clinical scanners. The distribution of radioactive tracers in small biological samples is commonly determined by film autoradiography. In this procedure, sample sections are usually applied directly to photographic film and one simply waits, perhaps weeks, until there have been enough interactions in the film to develop a satisfactory image. It has been suggested that image intensifiers offer a means for either improving the speed of image formation or for reducing the level of radioactivity required. Professor Reynolds at Princeton is exploring this idea in connection with general investigations of image intensifier applications in microscopy. In a technique that he has used, samples are applied to a 10 thick scintillator. The scintillations are then imaged by a lens system onto the photocathode of an image intensifier tube. The image tube output is bright enough to photograph in one second activity densities of 10-4 Ci distributed over a 10 x 10 sample area.

7

In the tracer studies and in other processes such as single cell bioluminescence there is a restricted amount of light available. The incorporation of image intensifiers has greatly improved observations in these limited illumination situations. Amplified brightness, however, is not always a complete solution for autoradiography or for microscopic observations in general, because there may still be insufficient information content in the images.

Increased speed of image formation has been especially useful in the study of bioluminescence dynamics. Good examples are pictures taken from the self luminescence of a single cell organism Noctiluca miliaris. This organism is shown in a 17 field of view in Fig. 4. The individual luminescing sites are clearly resolved and are apparently on the cell surface.

In collaboration with Buck and Hanson, National Institutes of Health, Reynold's image intensification system is being used to study the fine structure of a firefly flash. The light organ of a firefly is known from histological studies to be composed of many subunits, which

in turn contain many light-producing cells. Image intensifier system movies of the flash show that spots of light can be resolved during the flash and that these spot flashes are considerably shorter than the composite flashes, which last about 200 milliseconds. The measurements indicate that the subunit clusters of cells operate separately to make up a firefly flash. Movies taken at considerably higher magnificacions are to be attempted in the hope of resolving the responses of individual cells.

WHOLE BODY COUNTING

Whole body counters have been developed to measure and identify radioactivity in the body. Their great sensitivity makes it possible to measure natural body levels of radioactivity as well as activity acquired accidentally, such as from radioactive fallout or nuclear accidents. These counters have already proven their value in radiological health work and their great potential in medical diagnosis and research is just now 1 beginning to be realized.

The most widely used type of whole body counter was developed by Miller and Marinelli at Argonne National Laboratory. Their detector is a 4 x 8 in. sodium iodide crystal spectrometer inside a large heavy steel chamber. The subject being counted is seated inside the shield on a tilted chair or lies on the so-called "one-meter-arc couch.' The "tilted-chair" allows the subject partially to surround the detector at about 16 in. distance except for his lower limbs. This gives greater sensitivity than the "one-meter-arc couch," but the arc position is used whenever possible, because the total body goemetry is better.

A different type of whole body counter developed by Anderson at Los Alamos Scientific Laboratory utilizes a cylindrical shell filled with a liquid scintillant. The subject is placed inside the cylinder and thus is surrounded by the detector. An array of 24 in. diameter photomultiplier tubes monitor the liquid scintillant--an arrangement that has an obvious geometrical advantage.

The increased efficiency permits measurements to be made in a much shorter time than with the crystal system. Energy resolution is of course much better with the crystal detectors. Sharp photo-peak resolution is an important advantage in diagnoses and in general for the measurement and identification of body radioactivity.

Whole body counters are useful in medical studies such as iron absorption and excretion and calcium metabolism. They have been especially suitable for investigations of the abnormal body iron losses that can lead to iron deficiency anemia. Normally the percentage absorption of iron from the diet is not high and it is retained in the body and excreted slowly. A group study of such normal subjects given 59Fe has provided a standard with which to compare deviations in absorption and retention that are observed in a number of disease states.

In pernicious anemia an intrinsic factor that is necessary for Vitamin B12 absorption is no longer secreted by the intestinal mucosa. Simultaneous whole body counter retention measurements of 58Co labelled free Vitamin B12 and 60Co labelled Vitamin B12 bound to intrinsic factor can clearly distinguish normal persons from those with pernicious anemia and those with intestinal malabsorption.

47Ca studies have shown measurable differences in calcium metabolism between normal persons and patients with such problems as acromegaly and metastatic bone cancer.

Since potassium is widely distributed in the body except in fatty tissue, 40K can be used in determinations of lean body mass. Whole body 40K measurements have also suggested that a lowered potassium content is present prior to the onset and course of muscular distrophy.

IN VIVO STUDIES

Studies can also be carried on inside living systems. They usually require measurements that are restricted to small volume elements and are best performed with small semiconductor detectors. One mm diameter developmental probes are available and we expect to see even smaller ones available in the future. Small size is important for precise volume resolution and to minimize tissue damage and interference with organ function.

It is generally thought that blood flow within different parts of the cerebral tissue is regulated by the metabolism of the tissue itself. Measurements of blood flow in specific areas of the brain should, therefore, give information about the metabolism of the part studied. Some knowledge of regional blood flow has come from studies of cerebral structures that are accessible to detectors at the surface. A method that has

been used by Ingvar and Lassen in Norway involves monitoring the brain tissue clearance of 133Xe or 85Kr with external counters." 8 Following injection into the carotid artery one records either 133Xe gamma activity from a portion of the brain or 85Kr beta activity from superficial layers of the cortex. Since a number of important neurological diseases are caused by subsurface disturbances, more refined methods for quantitative measurements in the depths of the brain are necessary.

Beta sensitive probes have proved useful in brain tumor surgery. Their use is based on the observation that 32P concentrates relatively more in active brain tumors than in adjacent normal tissue. Because of this, a beta probe inserted into the tumor site can indicate the boundaries of the tumor as the probe passes through it. Accurate tumor location is especially helpful where it is necessary to remove a sample of tumor material for microscopic study.

Figure 5 shows an X-ray photograph taken during biopsy on a patient with a malignant glioma of the brain. The biopsy specimen was taken at a 5 cm. depth after final tumor boundary definition by the silicon junction detector probe. The photograph was furnished by Professor R. W. Ran University of California Medical School, Los Angeles.

Geiger tubes have been used in vivo by Bacaner at the University of Minnesota for investigations of blood flow variations in the intestines. In collaboration with Goulding, Lawrence Radiation Laboratory, Berkeley, semiconductor detectors have been developed that can be swallowed for studies of scomach and small bowel circulation. In the gastrointestinal tract of intact humans and animals circulation has been especially difficult to measure because of the length and irregular topography of the intestine. Consequently, we have had almost no previous information relating local blood flow changes in the bowel to normal functional variations and none at all relating to disfunctions and diseased states.

Heart studies in the dog have shown that blood flow in the tissue of the heart is quite inhomogenous and that this irregular flow distribution appears to be a critical determinant of heart performance. When coronary blood flow is increased in marginally contracting regions stronger muscular contractions result. Tracer studies of this mechanism have previously been made with geiger tubes applied to an isolated

heart. Goulding and Bacaner have also developed semiconductor detectors that can be fixed in arrays on the surface of a functioning dog heart for simultaneous direct blood flow measurements from several regions of the heart, Fig. 6.

LOW ENERGY MEASUREMENTS

There are problems with internal emitters such as 239Pu and 55Fe that require low energy X and gamma ray measurements. Scintillation spectrometry can be done down to about 10 Kev with special care for background and noise suppression. Proportional counters give better energy resolution and they are usable at somewhat lower energies than scintillation counters, but with about half the sensitivity. Both systems are useful for sample counting. Scintillation systems are generally used in low-energy whole body counting because they are more sensitive than proportional counters and routine use is easier.

The measurement of 55Fe which emits a 5.9 KeV X-ray is necessary because it is a widely distributed environmental contaminant. It should also prove useful as a biological tracer.

239Pu inhalation or wound contamination is a continuing significant potential problem in part of the nuclear industry. The diagnosis of quantities of 239Pu present in the body is difficult, because the photon emission that is relatively abundant is well below 20 KeV. Occasionally small amounts of Pu must be left in a wound because the counters cannot accurately localize the particles and the removal of significant amounts of additional tissue would cause physical impairment. This frequently occurs with hand wounds. Because of this there is a definite need for accurate means of locating concentrations of the order of 0.01 Ci so they can be removed.