CT stands for Computed Tomography, a method that measures X-ray attenuation from many directions to reconstruct cross-sectional images of the body. A plain X-ray shows mainly hard tissues like bone, but CT uses mathematical calculations to distinguish soft tissues as well. Hounsfield, working at EMI in the United Kingdom, built the first practical scanner and dramatically reduced the time needed to diagnose brain tumors or hemorrhages. Cormack independently formulated the theory for recovering density distributions from X-ray attenuation, an inverse Radon transform problem, providing the mathematical backbone. The technique lessened the need for invasive exploratory surgeries and improved survival rates in emergency medicine. Today, CT data can be fed into 3-D printers or VR systems for surgical planning and for teaching purposes. Thus, CT has become a revolutionary tool that fundamentally changed clinical practice.

The principle of CT is to rotate an X-ray source and detector array around the object and solve for the density function µ(x,y) through an inverse Radon transform of the collected projection data. The first EMI scanner acquired 180 projections at 1.5-degree increments and reconstructed an 80×80 matrix with 13-mm slice thickness in roughly five minutes. Hounsfield integrated NaI scintillation detectors with gamma correction electronics and applied a recursive iterative algorithm to suppress noise. Cormack, on the other hand, provided analytical solutions that considered weighted line integrals, allowing error estimation even when angular coverage was limited. These contributions were later standardized as the filtered back projection (FBP) algorithm, which underlies modern spiral and multislice CT systems. The CT number, or Hounsfield Unit, defines water as 0 and air as −1000, giving a quantitative scale for electron density differences among tissues. This quantitative nature enables applications ranging from bone mineral density assessment and emphysema grading to precise dose calculations in radiation therapy planning. With contrast media, dynamic CT can analyze organ perfusion and vascular function, and photon-counting detectors are pushing spectral CT from research labs into clinical practice. All these innovations trace back to the foundational theory and first-generation scanner recognized by the 1979 Nobel Prize.

The EMI Mark I scanner employed a translate-rotate geometry with four detector channels and produced an effective energy around 80 keV with a half-value layer of 1.8 mm Al. Beam hardening correction was carried out through an iterative refinement scheme that reduced streak artifacts while keeping patient dose acceptable. Cormack tackled limited-angle projection problems using Legendre polynomial expansions, deriving uniqueness conditions for Radon measurements and evaluating numerical stability. His framework was later generalized in Kak & Slaney’s classic tomographic text and extended to cone-beam CT via the Feldkamp algorithm. The Hounsfield Unit normalizes the linear X-ray attenuation coefficient by water, presupposing calibration that accounts for energy dependence and non-linearity. The first-generation scanner’s parallel beam, translate-rotate acquisition required more than 5 minutes per slice but achieved high contrast resolution of better than 2 HU. In the early 1980s, 2-D resampled FBP combined with sub-second gantry rotation and fan-beam geometry radically improved real-time imaging. Low-frequency gap compensation in Radon space and optimized weighting filters became precursors to today’s adaptive statistical iterative reconstruction techniques such as ASIR and MBIR. These advancements enabled a rigorous trade-off analysis between radiation dose reduction and image quality, leading to dose-length-product optimization based on diagnostic reference levels. The 1979 award remains a canonical example of synergy between mathematical theory and engineering implementation that propelled medical imaging forward.