Computed Tomography Imaging

Computed tomography (CT) reconstructs cross-sectional images of the body from many X-ray projections acquired around the patient. By solving for the attenuation at each point in a slice, CT overcomes the superimposition of projection radiography and displays anatomy as tomographic sections, with attenuation expressed on the Hounsfield scale. It is a workhorse for cross-sectional anatomical imaging.

Definition

Computed tomography is a cross-sectional imaging technique that reconstructs a map of X-ray attenuation within a slice of the body from projections acquired at many angles, expressing attenuation in Hounsfield units relative to water.

Scope

The topic covers the principle of reconstructing attenuation from multiple angular projections, the Hounsfield unit scale for quantifying tissue attenuation, the move from single-slice to helical and multidetector acquisition, and the radiation-dose considerations that accompany an ionising-radiation modality. It is a reference on how CT generates anatomical images, not clinical guidance.

Core questions

• How are cross-sectional attenuation maps reconstructed from angular projections?
• What does the Hounsfield unit scale represent and how does it relate to tissue type?
• How did helical and multidetector acquisition change the speed and coverage of CT?
• How is radiation dose balanced against image quality in CT?

Key concepts

• Tomographic reconstruction from projections
• Hounsfield unit (CT number) scale
• Windowing of attenuation values
• Helical and multidetector acquisition
• Isotropic voxels and multiplanar reformatting
• Radiation dose and dose optimisation

Mechanisms

An X-ray tube and detector array rotate around the patient, acquiring transmission measurements at many angles; a reconstruction algorithm solves these projections for the attenuation coefficient at each point in the slice, producing a cross-sectional image free of superimposition (Hounsfield, 1973). Attenuation values are normalised to the Hounsfield scale, where water is defined as 0 and air as approximately -1000, allowing tissue characterisation by CT number. Windowing maps a chosen range of these values to display grey levels. Helical (spiral) and multidetector acquisition let the scanner cover volumes rapidly and generate near-isotropic voxels suitable for multiplanar and three-dimensional reformatting. The physical and engineering basis is detailed in standard texts (Bushberg et al., 2012).

Clinical relevance

CT provides high-resolution cross-sectional depiction of skeletal, thoracic, abdominal, and vascular anatomy, and standardised descriptive terminology supports its consistent reading (Hansell et al., 2008). The first clinical demonstrations established its anatomical value (Ambrose, 1973). This entry describes how CT depicts anatomy and is not a basis for individual diagnostic or treatment decisions.

Epidemiology

CT delivers substantially higher radiation doses than conventional radiography and has become a major and increasing contributor to medical radiation exposure in many populations, which motivates careful justification and dose optimisation (Brenner & Hall, 2007).

History

Computed tomography was introduced by Godfrey Hounsfield, who described the system in 1973, with the mathematical foundations independently developed by Allan Cormack; James Ambrose reported the first clinical applications in the same year (Hounsfield, 1973; Ambrose, 1973). Hounsfield and Cormack shared the 1979 Nobel Prize in Physiology or Medicine. The modality evolved through single-slice, helical, and multidetector designs, progressively increasing speed, coverage, and reconstruction flexibility.

Key figures

• Godfrey Hounsfield
• Allan Cormack
• James Ambrose

Seminal works

• hounsfield-1973
• ambrose-1973

Frequently asked questions

What is a Hounsfield unit?

A Hounsfield unit is the normalised measure of X-ray attenuation used in CT, with water set to 0 and air to about -1000; different tissues occupy characteristic ranges, which helps identify them on the image.

How does CT differ from a plain radiograph?

A radiograph is a single projection in which structures are superimposed, whereas CT reconstructs cross-sectional slices from many projections, removing superimposition and allowing measurement of attenuation at each point.