The Examination of Internal Tissues by High-Energy Scattered X Radiation

The present technic is an extension to a much larger scale of research work at Hammersmith Hospital during 1958 and 1959 (2). The program has been supported by the British Empire Cancer Campaign over the last five years, during which an apparatus has been constructed and a series of experiments carried out in conjunction with a 5.6 MeV linear accelerator (1.a.). The x-ray output of the 1.a. is collimated to a narrow beam (3-mm diameter), which is passed through a patient. Rays scattered from a small volume of tissue (v in Fig. 1) are accepted by a large focusing collimator behind the patient and reach a liquid scintillator tank. The direct beam is absorbed, together with rays scattered either anteriorly or posteriorly to v. If the geometrical factors are held constant, the intensity of the scattered radiation reaching the scintillator depends on (a) the energy and intensity of the incident beam; (b) the absorption of the primary beam between the patient's anterior surface and v; (c) the absorption of the scattered radiation between v and the posterior surface of the patient; (d) the density (strictly electron density) of the tissue enclosed by v. Now (a) may be kept constant, and both (b) and (c) may be reduced by the use of high-energy x rays. Further, an approximate compensation for variations in absorption may be applied. The light output from the scintillator can therefore be made to depend primarily on the density of the tissue enclosed by v. The x-ray beam and collimator are arranged to scan the patient, while light from the scintillator modulates the brightness of a cathode-ray tube spot, whose movement corresponds with that of the x-ray beam, thus building up a picture of the density variations in the focal plane. The method is capable of resolving very small differences in density with low-exposure dose, and the ultimate aim is to detect differences of 1 per cent or less in a focal plane about 1 cm thick, with a linear resolution of 1 mm, by scans lasting a few seconds. The present paper reports the second stage in such a development and indicates what is needed for a third stage. R. L. Clarke (1) has used the basic principle in a modified form. The present paper is in substantial agreement on points common to his work. Description of the Apparatus In order to provide scans up to a rate of 30 lines per second, reciprocating motion was avoided and a system of continuously rotating collimators was used, as illustrated in Figure 2. Rays from the l.a. are first formed into a fan-shaped beam and then pass through one of the ten narrow apertures of the beam-defining wheel, forming a pencil beam which sweeps across in the direction of rotation.