Written and maintained by CASRAI Editorial Board
Last updated
Medical imaging is the broad scientific and technological field concerned with producing, interpreting, and analyzing visual representations of the interior of the human body — or of biological tissue more generally — for diagnosis, treatment guidance, and research. It is not limited to a single technique: X-rays, computed tomography (CT), magnetic resonance imaging (MRI), ultrasound, nuclear medicine (PET, SPECT), optical and photonic techniques, and whole-slide digital scanning of tissue specimens are all part of medical imaging, each turning a different physical signal — X-ray attenuation, magnetic resonance, sound reflection, radioactive decay, or light-tissue interaction — into an interpretable image. Medical imaging is the parent field to radiology, which is the clinical medical specialty organized around dedicated radiology and nuclear medicine departments and residency training; medical imaging as a research field is broader, also encompassing imaging that happens outside radiology departments entirely, such as digital pathology (scanning and analyzing microscope slides of tissue), ophthalmic imaging of the eye, and preclinical imaging of laboratory animals used to develop and validate a technique before it ever reaches a patient. Because every imaging modality is, at its core, an application of physical principles to biological tissue, medical imaging research draws directly and continuously on physics — the physics of X-ray attenuation and scatter, nuclear magnetic resonance, acoustic wave propagation, and light absorption/scattering all underlie how an imaging device actually produces a signal worth interpreting.
Medical imaging research centers on a recurring set of questions: What physical or molecular signal is being measured, and how faithfully can it be reconstructed into an image? How much radiation dose, contrast agent, scan time, or tissue disruption is required to get a useful result, and can that be reduced without losing diagnostic or scientific value? Can a measurement extracted from an image — a size, a density, a metabolic rate, a tissue pattern — serve as a reliable biomarker of disease presence, severity, or treatment response? And, increasingly, can computational and machine-learning methods assist or automate image interpretation reliably enough to trust in practice? Answering these questions is inherently interdisciplinary, pulling together physicists, biomedical and optical engineers, computer scientists, biologists, and clinicians, which is why medical imaging research programs are frequently organized across multiple departments rather than housed in just one.
Major Sub-Disciplines Within Medical Imaging
Clinical diagnostic and interventional imaging (radiology) is the largest and most familiar branch of medical imaging — X-ray, CT, MRI, ultrasound, and image-guided procedures organized within hospital radiology departments; see CASRAI’s dedicated guide to radiology for that clinical specialty in depth. Nuclear and molecular imaging uses radioactive tracers and PET/SPECT scanners to image metabolic and physiological processes rather than pure anatomy, and overlaps substantially with radiology while also standing as its own research tradition rooted in nuclear physics and pharmacology. Digital pathology uses high-resolution whole-slide scanners to digitize microscope slides of tissue specimens, turning what was traditionally a manual microscopy task into a large, computationally analyzable image dataset — a fast-growing subfield that increasingly borrows DICOM-based data standards originally built for radiology. Ophthalmic imaging — optical coherence tomography (OCT) and fundus photography of the eye — is a distinct clinical imaging tradition that sits within ophthalmology rather than radiology. Preclinical (small-animal) imaging applies scaled-down versions of clinical modalities (micro-CT, micro-PET, optical and bioluminescence imaging) to laboratory animals, forming the translational bridge between a new imaging idea and a human clinical trial. Optical and photonic imaging covers light-based techniques — fluorescence imaging, photoacoustic imaging, near-infrared spectroscopy — used in both basic biomedical research and emerging clinical applications, and draws heavily on optics and photonics. Medical physics and image science is the applied-physics discipline underlying image formation, radiation dose optimization, and equipment quality assurance across every modality above. Imaging informatics is the discipline of storing, exchanging, and computationally analyzing the resulting image data at scale — increasingly central as machine-learning-based image analysis has become a major methodological focus in its own right.
How Medical Imaging Research Is Funded
In the United States, medical imaging research is funded predominantly through the National Institutes of Health (NIH), but — because imaging is a tool applied across nearly every disease area rather than owned by one — that funding is spread across several institutes instead of concentrated in a single one. The National Institute of Biomedical Imaging and Bioengineering (NIBIB) is the NIH institute built specifically around imaging technology and bioengineering development; its mission centers on improving health through technology development in imaging and bioengineering, and it funds both the underlying physics/engineering of new imaging methods and enabling technologies used across the rest of NIH. Disease-focused institutes are major imaging funders within their own domains on top of NIBIB: the National Cancer Institute (NCI) funds oncologic imaging including cancer screening and detection research; the National Heart, Lung, and Blood Institute (NHLBI) funds cardiac and pulmonary imaging; the National Institute of Neurological Disorders and Stroke (NINDS) funds neuroimaging for stroke and neurological disease; the National Institute on Aging (NIA) funds imaging-biomarker research in dementia and Alzheimer’s disease; and the National Eye Institute (NEI) funds ophthalmic imaging research, including much of the foundational and clinical work behind OCT. At the National Science Foundation, imaging-related instrumentation, signal-processing, and optical/photonic imaging research — as distinct from the clinical, disease-focused imaging research NIH funds — is more likely to sit within the Directorate for Engineering, which supports biomedical engineering and imaging-instrumentation research more broadly.
Outside federal funding, the field’s own professional societies are a genuinely notable funding presence, particularly for early-career researchers: the Radiological Society of North America (RSNA) operates a research and education funding arm supporting imaging research grants, and the American College of Radiology (ACR), through its Neiman Health Policy Institute, funds and conducts health-services research on imaging utilization, cost, and policy — work that sits at the intersection of medical imaging and health economics.
Research Methods and Tools
The imaging systems themselves — radiography, CT, MRI, ultrasound, PET/SPECT scanners, optical and photoacoustic imaging systems, and whole-slide pathology scanners — are both the field’s clinical/research tools and its primary research instruments; a large share of imaging research develops new acquisition sequences, contrast or tracer agents, or reconstruction algorithms for these same classes of machine rather than building entirely new hardware. Images and their metadata are stored and exchanged using the DICOM standard and managed through a radiology information system (RIS) and picture archiving and communication system (PACS), often alongside a vendor-neutral archive for long-term, format-independent storage; documented DICOM conformance is what makes it possible to pool imaging data across institutions and vendors for multi-site research. Phantom studies — scanning a physical or digital object with known, controlled properties — are a standard way to validate a new imaging protocol’s accuracy and reproducibility before it is used on human subjects. Before a new modality or biomarker reaches human trials, it is typically validated first in preclinical animal models using scaled-down imaging equipment. Quantitative imaging biomarker validation generally then requires prospective clinical studies comparing an imaging-derived measurement against a clinical outcome or an established reference standard. Image analysis increasingly involves computational methods — from manual expert annotation, through semi-automated segmentation, to machine-learning models trained to detect or classify findings — and rigorously evaluating those models on data they were not trained on, ideally drawn from multiple institutions, is itself now a significant methodological focus across the field.
Career and Training Pathways
Clinical imaging careers branch by specialty: physicians who read and perform diagnostic/interventional imaging in a hospital setting train through an MD or DO degree followed by a radiology residency and board certification through the American Board of Radiology (ABR), while ophthalmic imaging sits within ophthalmology training and digital pathology sits within pathology training, each with its own residency pathway. Non-physician imaging scientists — medical physicists and biomedical/optical/imaging engineers — typically hold a PhD in medical physics, biomedical engineering, optics/photonics, or a related field grounded in physics; medical physicists who work clinically (for example on radiation dose and equipment quality assurance) generally also complete board certification. Relevant professional societies span the field’s clinical and technical halves: RSNA and ACR for the clinical/diagnostic side; the American Association of Physicists in Medicine (AAPM) for medical physicists; the Society for Imaging Informatics in Medicine (SIIM) for the informatics side (PACS, DICOM, image data management); the International Society for Magnetic Resonance in Medicine (ISMRM) for MRI-focused researchers; and SPIE (the international optics and photonics society) for researchers working in optical and photonic imaging. Researchers focused on the policy and economics side of imaging — utilization, cost-effectiveness, appropriate-use criteria — often train through health-services research programs adjacent to health economics rather than through a clinical or engineering PhD track.
Related Disciplines
Medical imaging’s research questions connect closely to several neighboring fields. Its clinical core is radiology, and its underlying physical principles come directly from physics. Preclinical imaging depends on animal science for the laboratory-animal models a new technique is validated in before human use. The policy and utilization side of the field overlaps substantially with health economics, particularly in research on imaging cost-effectiveness and appropriate use. And at the molecular scale, structural-imaging techniques such as X-ray crystallography and cryo-electron microscopy — covered in CASRAI’s guide to crystallography — apply closely related imaging principles to determine the three-dimensional structure of individual biomolecules, several orders of magnitude smaller than what a clinical scanner resolves, but built on the same underlying idea of reconstructing structure from a measured physical signal.
Explore More Branches of Science
This guide is part of CASRAI’s Branches of Science hub, which organizes and cross-links guides to individual scientific and academic disciplines by category — Physical Sciences, Life Sciences, Formal Sciences, Social Sciences, and Applied Sciences & Engineering — each written with the research-administration depth (funding landscape, methods, career pathways) that generic encyclopedia overviews leave out. Visit the hub to find guides to related disciplines, including radiology, physics, and other applied-science and life-science fields.
Frequently Asked Questions
What is the difference between medical imaging and radiology?
Radiology is the clinical medical specialty organized around hospital radiology and nuclear medicine departments — X-ray, CT, MRI, ultrasound, and image-guided procedures read and performed by physician radiologists. Medical imaging is the broader field: it includes radiology but also imaging that happens outside radiology departments entirely, such as digital pathology, ophthalmic imaging, and preclinical imaging in laboratory animals.
What NIH institute funds medical imaging research?
No single institute owns all of medical imaging the way a disease-focused institute owns its disease area. The National Institute of Biomedical Imaging and Bioengineering (NIBIB) is the institute built specifically around imaging technology development, but because imaging is applied across nearly every disease area, institutes such as NCI, NHLBI, NINDS, NIA, and NEI each fund substantial imaging research within their own domains.
Is medical imaging only about clinical diagnosis, or does it include basic research?
Both. Clinical imaging research evaluates how a technique performs in diagnosing or monitoring disease in patients. A substantial parallel track of imaging physics, optics, and engineering research develops the new acquisition methods, reconstruction algorithms, contrast agents, and hardware that clinical imaging eventually uses — work that is closer to basic and applied physical science and is typically validated first in preclinical animal models before any human study.
What is imaging informatics?
Imaging informatics is the subfield concerned with storing, exchanging, and computationally analyzing medical images at scale — the DICOM data standard, PACS and vendor-neutral archive infrastructure, and increasingly the machine-learning pipelines used to assist image interpretation. It sits at the intersection of medical imaging and health information technology.
Do I need to be a physician to have a career in medical imaging research?
No. While physician radiologists, ophthalmologists, and pathologists all do clinical imaging research, a large share of medical imaging research is conducted by non-physician scientists — medical physicists, biomedical engineers, and optical/imaging engineers — who typically hold a PhD rather than a medical degree, and whose work focuses on the underlying technology rather than direct patient care.
