The History of Transverse Axial Tomography prior to CT

Introduction

All internet searches of the term transverse axial tomography yield results that are expressed in terms of Computed Tomography (CT). This technique came into clinical practice in 1970 following Godfrey Hounsfield’s brilliant technical development. However, analogue methods for producing transverse axial tomographic images originated in the 1930’s under the technique called horizontal body section radiography. The first equipment for this purpose was built in 1937 in the UK (possibly the world) by W. Watson and Sons, a well-known UK X-ray equipment manufacturer. British and US patents being issued to the company in 1937.  A history of the company can be found on Wikipedia and during this period the UK had a healthy X-ray and related manufacturing industry covering most aspects of imaging and radiotherapy equipment.

Although methods for producing transverse section images originated in the 1930’s, clinical interest in cross-sectional anatomy dates from 1875 with the production of a “Topographic anatomical atlas according sections on frozen cadavers, by Braune, second edition, Leipzig: Veit and Co. 1875.” Several other atlases were produced culminating in the production of an “Atlas of Cross-Sections of the Body for Use in Roentgen Deep Therapy, produced by Hans Holfelder and published by Julius Springer, Berlin in 1924.

The Atlas reproduced some 38 body sections with 4 of the head and 34 from the shoulders to the lower abdomen. As indicated in the title of the Atlas, its purpose was to create improved treatment plans by:

“Using the cross-section diagrams in which clinical findings have been inserted together with field selection and the knowledge of the favourite ways of the topographical extension of the disease and by considering the relative location of each organ we have been able to obtain much better and more adequate results in the practical application of the roentgen deep therapy than ever before.”

Obviously, the cross sections employed were not patient specific, but they did provide an anatomical background that could be used for treatment planning. An example of treatment plans overlaid on a transverse anatomical section is shown in Figure 1.

Development of trans axial tomography equipment meant that patient specific cross- sectional anatomy could be employed in treatment planning. A detailed review of the application of this new type of equipment in clinical practice was published in the British Journal of Radiology in 1950 in a comprehensive paper by J.J. Stevenson entitled Horizontal Body Section Radiography (Vol XX11, No 270, pages 319 – 334).

Carcinoma

Figure 1.   Radiation of a carcinoma of the lesser curvature of the stomach with four fields arranged in an obliques section. The suparenals are avoided by the obliques section and the spleen by the arrangement of the fields.

Although invented in the UK the technique did not find favour in Britain. However, Dr M. J. Ryan, a Senior Registrar at the time at the Christie Hospital and Holt Radium Institute returned from the Third International Congress on Stratigraphy held in Genoa in 1955 and reported on apparatus for transverse axial tomography which had been exhibited at the Congress.

Because rotation therapy was in vogue at the Christie Hospital it was decided to explore the possibilities of this form of imaging both from a diagnostic as well as a therapeutic perspective. Furthermore, it was decided to explore its possible application to the demonstration of malignant disease in the mediastinum, a region that was extremely difficult to investigate with conventional methods. Consequently, it  a Zuder “Assistrator” tomographic unit was purchased and a team was established to investigate its clinical applications.

Trans axial tomographic equipment

The arrangement of a transverse axial tomographic equipment is shown in Figure 2. The patient was seated on a rotating saddle close to a horizontal cassette and the X-ray beam was directed at an angle of 20-30° below the horizontal. During an exposure the saddle and the cassette rotate in synchrony at the same uniform speed. The X-ray beam is collimated by means of an elliptically shaped diaphragm, which helps to reduce scattered radiation and improves radiographic quality.  The resulting radiograph is a tomographic section parallel to the cassette. This is true if certain conditions are met and the resulting radiograph is a classic example of an analogue back projection process as the transverse X-ray beam is smeared across the horizontal cassette.

Transverse Axial Tomography

Figure 2.        General layout of the equipment

Important considerations for effective image production were:

  1. The axes of rotation of both the patient saddle and film cassette must be parallel and vertical.
  2. The patient and film should rotate smoothly and in synchrony.
  3. The lateral position of the X-ray tube focal spot should be accurate to within 0.5 mm of the line joining the focal spot and the two vertical axes of rotation of the saddle and film cassette.
  4. To know the level of cut to within ± 2 mm requires the vertical position of the focal spot to be set within roughly ± 8 mm.
  5. Provision for raising and lowering both the cassette and saddle in order to produce sections at different levels; and a device to indicate the level of cut.

The cut height could be indicated by means of an optical projector that was focussed to produce a cross of light at any convenient distance. This moves up and down a vertical column shown to the right of Figure 2.

Quality Assurance and Control

Because the X-ray tube employed for trans axial tomography was also used for other radiographic applications it was necessary to undertake a number of tests on the equipment prior to its tomographic use. Blurring was an important consideration and both geometric, due to the focal spot size, and screen film blurring were present. For the system employed these were assessed to be acceptable. However, of some importance was the blurring that could be introduced by incorrect alignment or lack of synchrony in the movement of the saddle and the cassette holder. To monitor this a vertical test rod of approximately 6 mm diameter was positioned vertically at the centre of the saddle and a tomographic image produced. Figure 3 shows the image of the rod when there was an error in synchronisation and when the error had been eliminated.

Transverse Axial Tomogram

Figure 3.  Synthesis of a transverse axial tomogram from a succession of equally spaced stationary exposures (a) when there is an error of synchrony; (b) when there is perfect synchrony.

Note: thanks to the wonders of modern technology, the original image has been upscaled using Artificial Intelligence.

In order to enable tests to be performed that corresponded to higher levels of the thorax, a test assembly comprising thirteen metal rods, each of 13 mm diameter and 1 metre long held apart by five Perspex plates so as to form a cylinder, was employed. This could be clamped in place of the saddle 4 resulting in an image of a section through the rods forming a circular pattern.

The equipment was in routine use throughout the 1970’s and a second unit located in Sheffield was transferred to the Christie Hospital as a possible source of spare parts. However, the purchase by the Christie Hospital of an EMI 7070 body section CT scanner in 1980 rendered the analogue technique redundant.

 

Mike Moores

Director