What Is Tomography?

Tomography is an imaging method that reconstructs the internal structure of an object from a series of measurements taken at multiple angles or positions around it, without requiring physical sectioning or penetration. The word derives from the Greek tomos (a slice) and graphia (writing), reflecting the technique's goal of producing two-dimensional cross-sectional images that, when stacked, describe a three-dimensional volume. A tomographic measurement system records how a probe signal, such as an X-ray beam, magnetic field, sound wave, or electrical current, is attenuated, scattered, or altered by the material it passes through. Mathematical reconstruction algorithms, particularly the filtered back-projection method and iterative algebraic reconstruction techniques, then invert these measurements to produce the image. Tomography draws from physics, mathematics, signal processing, and electrical engineering, and the 1979 Nobel Prize in Physiology or Medicine was awarded jointly to Godfrey Hounsfield and Allan Cormack for developing computed tomography.

Computed Tomography

Computed tomography (CT) uses a rotating X-ray source and an arc of detectors to acquire hundreds of projection images as the apparatus circles around a patient or object. Each rotation provides a set of line-integral measurements of X-ray attenuation through the cross-section; the filtered back-projection algorithm reconstructs a 2D slice image from each rotation, and consecutive slices are stacked to form a 3D volume. Modern multidetector CT systems acquire 64 or more slices per rotation, enabling whole-body imaging in seconds. The National Institute of Biomedical Imaging and Bioengineering describes how a standard clinical CT scanner delivers spatial resolutions of under one millimeter, sufficient to visualize coronary arteries, pulmonary nodules, and bone microstructure. Photon-counting CT detectors, first approved for clinical use in 2021, convert X-ray photons directly to electrical signals without the intermediate light-conversion step of conventional detectors, improving energy resolution and spatial performance.

Other Tomographic Modalities

Magnetic resonance imaging (MRI) is a tomographic technique that uses radiofrequency excitation of hydrogen nuclei in a strong magnetic field; spatial encoding by gradient fields allows reconstruction of anatomical slices without ionizing radiation, with superior soft-tissue contrast compared to CT. Positron emission tomography (PET) reconstructs the distribution of a radioactive tracer from coincidence detection of the annihilation photons it emits, giving functional and metabolic information rather than purely anatomical structure. Electrical impedance tomography (EIT) reconstructs the internal conductivity distribution of a body from surface electrode measurements, offering a radiation-free bedside monitoring option. Industrial applications use X-ray CT, neutron tomography, and ultrasonic computed tomography for non-destructive inspection of manufactured parts, castings, and composite structures. Research on all these modalities is published regularly in IEEE Transactions on Medical Imaging.

Image Reconstruction

The mathematical problem at the core of tomography is an inverse problem: given a set of projections, recover the spatial function that produced them. Filtered back-projection provides an analytic solution for the idealized parallel-beam case and remains computationally efficient, but it amplifies noise and produces artifacts when projections are sparse or incomplete. Iterative reconstruction methods, including algebraic reconstruction technique (ART) and statistical methods based on maximum likelihood expectation maximization (MLEM), model the measurement process explicitly and incorporate prior information about image properties, enabling sharper images at lower radiation doses. Deep learning reconstruction, where neural networks replace or supplement traditional algorithms, has been an active area since approximately 2016, with arXiv preprints and journal papers describing networks trained to reduce CT artifacts from sparse-view acquisitions.

Applications

Tomography has applications in a wide range of fields, including:

• Clinical diagnosis and staging of cancer, cardiovascular disease, and neurological conditions
• Industrial non-destructive testing of aerospace components, welds, and composite structures
• Geophysical seismic tomography for mapping subsurface rock formations
• Security screening in baggage and cargo inspection systems
• Scientific imaging in materials science, paleontology, and microelectronics failure analysis