How Medical Imaging Works: The Physics Behind the Pictures

Medical imaging translates invisible physical phenomena — electromagnetic radiation, sound waves, magnetic fields, and radioactive decay — into spatial maps of human anatomy and physiology. Understanding the underlying physics clarifies why different modalities answer different clinical questions, why some carry radiation dose while others do not, and how image quality is governed by measurable parameters rather than operator intuition. This page covers the core physical principles behind the major imaging technologies used in diagnostic and interventional radiology, explains how those principles create clinical images, and identifies where modality boundaries intersect with patient safety standards.

Definition and scope

Medical imaging encompasses any technique that produces a visual representation of internal body structures or function without surgical exposure. The field is broadly organized by the American College of Radiology (ACR) into modalities classified by their energy source: ionizing radiation (X-ray, CT, fluoroscopy, nuclear medicine, PET) and non-ionizing energy (MRI, ultrasound). The distinction matters clinically and regulatorily because ionizing radiation deposits energy in tissue in ways quantified by effective dose, measured in millisieverts (mSv), while non-ionizing modalities operate under different risk categories.

The U.S. Food and Drug Administration (FDA) regulates radiation-emitting imaging equipment under 21 CFR Part 1020, which establishes performance standards for diagnostic X-ray systems. The Nuclear Regulatory Commission (NRC) governs radioactive materials used in nuclear medicine and PET under 10 CFR Part 35. This regulatory framework is explored in greater depth at the regulatory context for radiology reference on this site.

How it works

Each modality exploits a distinct physical interaction between an energy source and biological tissue.

X-ray and computed tomography (CT)

Conventional X-ray imaging passes a collimated beam of photons — typically generated at 40–150 kilovolts peak (kVp) — through the body. Dense structures such as cortical bone attenuate more photons; air-filled structures attenuate fewer. The differential attenuation pattern is captured on a digital detector as a two-dimensional projection.

CT scanning extends this principle by rotating an X-ray tube and detector array 360 degrees around the patient, acquiring hundreds of projections. Reconstruction algorithms — most commonly filtered back projection or iterative reconstruction — synthesize these projections into volumetric datasets with sub-millimeter spatial resolution. A modern 64-slice CT scanner can acquire a chest-to-pelvis dataset in under 10 seconds. Effective dose for an abdominal CT is approximately 8 mSv, compared with roughly 0.1 mSv for a chest X-ray (ACR Dose Index Registry, reported methodology).

Magnetic resonance imaging (MRI)

MRI uses no ionizing radiation. Instead, it exploits the magnetic moment of hydrogen nuclei (protons) abundant in tissue water and fat. When placed in a strong static magnetic field — clinical scanners operate at 1.5 tesla (T) or 3.0 T — protons align with the field. Radiofrequency (RF) pulses delivered at the Larmor frequency knock protons into a higher energy state; as they relax back, they emit RF signals detected by receiver coils. Spatial encoding is achieved through magnetic field gradients that cause protons at different positions to precess at slightly different frequencies, allowing Fourier transform reconstruction of slice location and tissue contrast.

Different pulse sequences (T1-weighted, T2-weighted, FLAIR, diffusion-weighted) emphasize different relaxation properties, enabling discrimination of tissue types that appear identical on CT.

Ultrasound

Ultrasound transmits mechanical pressure waves — typically at frequencies between 2 and 15 megahertz (MHz) — from a piezoelectric transducer into tissue. At interfaces between tissues of different acoustic impedance, partial reflection occurs. The transducer detects returning echoes; time-of-flight calculations convert echo delay into depth, and amplitude maps spatial structure in real time. Because it uses no ionizing radiation and produces images in real time, ultrasound is the primary modality for imaging during pregnancy per ACR and American Institute of Ultrasound in Medicine (AIUM) guidelines.

Nuclear medicine and PET

Nuclear medicine and PET scanning differ fundamentally from anatomic imaging: they image physiologic function by tracking the biodistribution of radiolabeled compounds. In PET, a positron-emitting radiopharmaceutical — most commonly fluorine-18 fluorodeoxyglucose (FDG) — concentrates in metabolically active tissue. Positron annihilation produces paired 511-keV gamma photons detected in coincidence by the scanner ring. The resulting image reflects glucose metabolism rather than anatomy, which is why PET/CT fusion became standard for oncologic staging.

Common scenarios

The physics of each modality maps directly to clinical application:

  1. Suspected fracture or acute chest pathology — X-ray is the first-line tool because bone-to-soft-tissue contrast from photon attenuation differences is diagnostically adequate and radiation dose is minimal.
  2. Abdominal or pelvic pain, complex trauma — CT delivers sub-centimeter spatial resolution with rapid acquisition, critical when anatomy must be surveyed in a hemodynamically unstable patient.
  3. Brain, spine, or soft-tissue tumor characterization — MRI's superior soft-tissue contrast and multiplanar capability without radiation make it the standard modality for neurological and musculoskeletal questions addressed in imaging joint pain and musculoskeletal conditions.
  4. Cardiac or vascular structure in real time — Ultrasound's frame rate (up to 60 frames per second on modern systems) captures dynamic function that static modalities cannot.
  5. Metabolic activity in known malignancy — PET/CT detects lesions based on glucose uptake rather than anatomic size threshold, reducing false-negative staging compared to CT alone for certain tumor types.

Decision boundaries

Choosing among modalities requires balancing four competing variables: diagnostic yield, radiation dose, patient-specific contraindications, and resource availability. The ACR publishes Appropriateness Criteria — evidence-based guidelines rating each modality for over 200 clinical scenarios — which form the operational standard for radiologist consultation and ordering physician guidance. The overview of radiology practice contextualizes how these decisions fit within the broader clinical workflow.

Key contraindication boundaries include:

The physics of image formation are not separable from these decision boundaries — spatial resolution, signal-to-noise ratio, and acquisition time are all parameters that must be traded against biological risk and diagnostic necessity on a per-patient basis.

References

The law belongs to the people. Georgia v. Public.Resource.Org, 590 U.S. (2020)