Interventional radiology is a medical specialty that uses imaging to guide instruments inside the body for diagnosis and treatment. Its procedures typically reach a target through a small skin puncture, a blood vessel, or an existing anatomical passage rather than through a large surgical incision. Interventional radiologists use needles, guidewires, catheters, balloons, stents, embolic materials, and energy-delivery devices to sample tissue, restore or obstruct blood flow, drain collections, relieve obstruction, and destroy selected tumors. The specialty combines image interpretation with procedural technique and clinical management before and after an intervention.
Developed from diagnostic radiology during the twentieth century, interventional radiology now contributes to emergency hemorrhage control, cancer treatment, management of arterial and venous disease, and many forms of organ-preserving care. Its boundaries overlap with vascular surgery, gastroenterology, urology, cardiology, and neurosurgery; the allocation of particular procedures varies between institutions and countries. “Minimally invasive” describes the route of access, not necessarily the biological impact or risk of treatment: a puncture-sized entry can support a major intervention requiring anesthesia, intensive monitoring, and prolonged follow-up.
Historical development and the emergence of a clinical specialty
The development of interventional radiology depended on techniques that made internal structures both visible and accessible. Early angiography demonstrated blood vessels by injecting radiopaque material and recording its distribution with X-rays. In 1927, the Portuguese physician Egas Moniz introduced cerebral angiography, establishing an important foundation for imaging the brain’s circulation. Early vascular examinations could require surgical exposure of an artery or direct puncture with relatively large needles, limiting their convenience and increasing procedural hazards.
A decisive advance came in 1953, when the Swedish radiologist Sven-Ivar Seldinger described a method of catheter placement using a needle, flexible guidewire, and catheter. In the Seldinger technique, the wire preserves access after the needle is removed, allowing a catheter to be introduced without surgical exposure of the vessel. This principle became fundamental not only to angiography but also to central venous access, drainage procedures, and many other image-guided interventions. Later refinements included smaller access systems, hydrophilic wires, shaped catheters, and ultrasound-guided puncture.
In 1964, Charles Dotter and Melvin Judkins reported percutaneous transluminal treatment of atherosclerotic arterial narrowing using coaxial dilating catheters. Their work demonstrated that a catheter could be a therapeutic instrument rather than merely a means of obtaining images. Andreas Grüntzig subsequently developed balloon-catheter techniques, performing peripheral balloon angioplasty in the 1970s and the first successful percutaneous coronary balloon angioplasty in 1977. Coronary intervention became predominantly a cardiology discipline, while many peripheral vascular procedures remained shared among several specialties.
Other developments broadened the field beyond reopening arteries. Catheter-directed embolization made it possible to deliberately occlude abnormal or bleeding vessels. Image-guided drainage offered an alternative to surgical drainage for many localized fluid collections. Improvements in ultrasound and computed tomography enabled increasingly accurate needle placement in organs and deep tissues. Expandable metallic stents, microcatheters, detachable coils, stent-grafts, and thermal ablation systems expanded the range of treatable conditions during the late twentieth century.
Professional organization followed technical growth. The Society of Cardiovascular Radiology, founded in the United States in 1973, later became the Society of Cardiovascular and Interventional Radiology and adopted the name Society of Interventional Radiology in 2002. Increasing emphasis on outpatient consultation, inpatient care, and longitudinal follow-up distinguished the modern specialty from a procedure-only service. Training structures nevertheless differ internationally: some systems recognize interventional radiology as a primary specialty or formal subspecialty, while others organize it within broader radiology training.
Imaging guidance, access, and procedural instruments
Interventional imaging must show both the target and the instrument approaching it. Ultrasonography provides real-time images without ionizing radiation and is particularly useful for vascular access, superficial biopsy, abdominal drainage, and selected ablation procedures. Doppler techniques add information about blood flow. Its limitations include acoustic obstruction by bone or gas, reduced penetration in some patients, and difficulty visualizing a needle when its orientation produces weak echoes. Maintaining visibility of the needle tip is more important than simply seeing part of the needle shaft.
Fluoroscopy produces a continuous or pulsed sequence of X-ray images, making it suitable for tracking radiopaque wires, catheters, and implanted devices. When a contrast agent is injected into a vessel or duct, its course and patency become visible. Digital subtraction angiography removes a pre-contrast background image from subsequent images, improving the conspicuity of contrast-filled vessels. Patient movement can create subtraction artifacts, and the technique provides a projection rather than an inherently three-dimensional representation.
Computed tomography provides cross-sectional localization of deep targets, including small lung nodules, retroperitoneal collections, and lesions near complex anatomy. Instruments may be advanced incrementally between scans or under CT fluoroscopy. Cone-beam CT, obtained with a rotating angiographic system, adds three-dimensional information within the procedure room and can help identify tumor-feeding vessels or assess device position. Magnetic resonance imaging offers strong soft-tissue contrast without ionizing radiation, but restricted access, longer acquisition times, and the requirement for compatible instruments limit its use in routine interventions.
The access system is assembled around the intended task. An introducer sheath creates a stable entry channel; a guidewire provides support; and a catheter carries contrast, drugs, or devices to the target. Microcatheters permit selective navigation into small branches and can pass through larger guiding catheters. Device diameter is commonly expressed in French units: corresponds to approximately . For catheters, this generally describes external diameter, whereas sheath labeling commonly refers to the size of catheter accommodated.
Image guidance also involves planning a safe path. Operators assess intervening vessels, bowel, pleura, nerves, and other structures before puncture. In selected procedures, fluid or gas can be introduced to separate a treatment target from a vulnerable adjacent organ. Image fusion can align preprocedural CT or MRI with live ultrasound or fluoroscopy, although respiratory movement and tissue deformation can reduce registration accuracy. Final imaging documents the immediate technical result, but its meaning depends on the intervention: an open vessel, a functioning drain, and an adequate ablation zone require different forms of confirmation.
Arterial revascularization, aneurysm repair, and hemorrhage control
Arterial interventions pursue two apparently opposite goals: restoring flow where it is inadequate and stopping flow where it is harmful. In peripheral arterial disease, atherosclerotic narrowing or occlusion can limit blood supply to a limb. Angioplasty expands a narrowed segment with an inflatable balloon, while a stent provides a supporting scaffold when appropriate. Recoil, dissection, calcification, thrombosis, and later tissue growth within the treated segment influence durability. Drug-coated balloons and drug-eluting stents deliver antiproliferative agents intended to reduce restenosis, but their role depends on vascular territory, lesion characteristics, and the evidence for the particular device.
An aneurysm may be treated by excluding it from arterial pressure rather than removing it. Endovascular aneurysm repair uses a fabric-covered stent framework to channel blood through the diseased segment. Anatomical suitability depends on factors such as vessel diameter, access-vessel size, and the length and shape of the sealing zones. A major follow-up issue is an endoleak: persistent flow outside the graft but within the aneurysm sac. Different endoleak mechanisms have different implications, and successful implantation does not eliminate the need for surveillance or possible reintervention.
Embolization deliberately blocks a vessel using coils, plugs, particles, liquid agents, or other materials. In traumatic injury, gastrointestinal hemorrhage, postoperative bleeding, and postpartum hemorrhage, selective catheterization can identify and occlude a bleeding branch while preserving as much normal circulation as possible. The appropriate material depends on vessel size, flow, collateral supply, and whether permanent or temporary occlusion is intended. Coils provide a framework for localized occlusion, whereas particles can reach smaller downstream vessels. Liquid embolic agents have distinct flow and solidification properties that require precise delivery.
The same method can reduce pathological blood flow in nonemergency settings. Uterine artery embolization treats symptomatic fibroids by reducing their arterial supply, usually causing infarction and subsequent shrinkage of fibroid tissue. Embolization also has roles in selected vascular malformations and in reducing blood loss before certain operations. Clinical outcomes are not captured by angiographic appearance alone: symptom relief, tissue preservation, need for later treatment, and effects on organ function are separate endpoints.
The cerebral circulation represents a highly specialized branch of endovascular practice. Aneurysm coiling and flow-diverting devices treat selected intracranial aneurysms, while mechanical thrombectomy removes occluding clot in eligible patients with acute ischemic stroke. Neurointerventional procedures are performed by physicians with dedicated training drawn from several specialties. Similarly, percutaneous coronary intervention shares many catheter principles with interventional radiology but is generally performed within interventional cardiology. These overlaps reflect a common technical foundation rather than a single uniform professional jurisdiction.
Venous interventions, dialysis access, and portal circulation
Venous interventions address obstruction, abnormal flow, difficult vascular access, and complications of chronic organ disease. Image-guided central venous catheter placement is used for prolonged intravenous treatment, implanted access ports, and selected forms of nutritional support. Ultrasound helps identify a patent vein and guide puncture, while fluoroscopy can establish catheter course and tip position. Mechanical placement is only one part of success: later infection, thrombosis, migration, fracture, or malfunction may require investigation and treatment.
In deep vein thrombosis, catheter-directed thrombolysis delivers a clot-dissolving drug into thrombus, while thrombectomy devices remove or disrupt clot mechanically. These approaches are not interchangeable with routine medical treatment for every patient. The 2017 ATTRACT trial found that adding pharmacomechanical catheter-directed thrombolysis to anticoagulation did not reduce the overall occurrence of post-thrombotic syndrome in the trial population and increased early major bleeding. Findings varied by outcome and anatomical subgroup, illustrating why thrombus location, symptom severity, bleeding risk, and the treatment objective matter when interpreting evidence.
Venous stents can support narrowed central or pelvic veins in selected obstructive disorders. Inferior vena cava filters are designed to intercept emboli traveling from the lower-body venous circulation toward the lungs. They do not dissolve existing clot or treat the underlying tendency to thrombosis. Their use is limited by complications including new thrombosis, penetration, migration, and device fracture. Retrievable designs create an additional clinical responsibility: whether a filter remains necessary and whether removal is feasible are distinct questions requiring follow-up.
Interventional radiology also supports hemodialysis access. Arteriovenous fistulas and grafts can develop stenosis, thrombosis, or abnormal flow that compromises dialysis or affects the limb and circulation. Fistulography maps the access circuit; angioplasty treats selected stenoses; and thrombectomy may restore a thrombosed access. These procedures are directed toward preserving a usable access over time, rather than simply producing a satisfactory image at the end of one intervention.
The transjugular intrahepatic portosystemic shunt, or TIPS, creates a channel through the liver between the portal and systemic venous circulations. A covered stent maintains the channel, reducing portal pressure in selected complications of cirrhosis, including variceal bleeding and refractory ascites. The physiological trade-off is important: blood bypasses some hepatic processing, potentially contributing to hepatic encephalopathy, while increased venous return can place additional demands on the heart. Portal vein embolization has a different purpose: occluding selected portal branches before liver resection can stimulate growth of the liver portion intended to remain.
Interventional oncology and local tumor treatment
Interventional oncology applies image-guided procedures to the diagnosis and treatment of cancer. It includes tissue acquisition, tumor ablation, intra-arterial therapies, and procedures that relieve complications of malignancy. Its role depends on tumor type, distribution, organ function, and the availability of surgery, systemic treatment, and radiation therapy. Treatment may aim at local eradication, disease control, preparation for transplantation or surgery, or symptom relief. These aims must be distinguished when comparing outcomes across studies.
Image-guided biopsy obtains tissue from a lesion while avoiding vulnerable structures. Fine-needle aspiration predominantly provides cells, whereas core biopsy obtains tissue architecture and often more material for histological and molecular testing. Adequacy depends on accurate targeting, specimen handling, and sampling of viable rather than necrotic tissue. A technically successful needle placement does not guarantee a diagnostic specimen. For some tumors, the biopsy route also matters because later surgery may need to encompass the tract or because traversing particular compartments can complicate subsequent treatment.
Tumor ablation destroys tissue in situ. Radiofrequency ablation generates heating through electrical current, microwave ablation deposits electromagnetic energy, and cryoablation uses cycles of freezing and thawing. The aim is usually to cover the visible tumor together with an appropriate surrounding margin. Nearby blood vessels can remove heat and reduce thermal injury to part of the target, a phenomenon often called the heat-sink effect. Conversely, heat or cold extending beyond the intended zone can injure bowel, bile ducts, skin, or nerves. Lesion size, location, imaging visibility, and proximity to critical structures therefore strongly affect feasibility.
Transarterial treatment exploits the arterial supply of certain tumors. In hepatocellular carcinoma, an important form of liver cancer, malignant tissue commonly receives a substantial proportion of its blood supply from the hepatic artery. Transarterial chemoembolization combines arterial delivery of cytotoxic treatment with embolization. Radioembolization delivers radioactive microspheres, commonly containing yttrium-90, into selected hepatic arteries. Its planning includes assessment of vascular anatomy, unintended extrahepatic delivery, and shunting toward the lungs. Delivered radiation dose and the distribution of activity are central to treatment planning.
Response assessment differs from simply measuring whether a mass has disappeared. An ablated tumor may remain visible as a nonenhancing treatment zone, while a successfully treated liver tumor may shrink slowly. Contrast enhancement, viable-tumor criteria, local progression, overall survival, and patient-reported symptoms provide different information. Interventional oncology is consequently integrated with multidisciplinary cancer care: local control of one lesion does not necessarily control microscopic or distant disease, and apparent imaging response is not automatically equivalent to a survival benefit.
Drainage, organ access, and musculoskeletal procedures
Many nonvascular interventions create a controlled route into an organ, duct, or abnormal collection. Percutaneous drainage is used for selected abscesses and other fluid collections in the abdomen, pelvis, chest, and soft tissues. Imaging establishes the collection’s extent, the safest access path, and whether contents are likely to pass through a catheter. Drainage can provide material for microbiological analysis and reduce an infected collection’s burden, but it does not automatically correct an underlying bowel leak, obstructed duct, or devitalized tissue. Such causes can determine whether drainage is definitive or one component of broader treatment.
Biliary intervention addresses obstruction or leakage within the bile ducts. Percutaneous transhepatic access passes through the liver into an intrahepatic duct, permitting drainage, dilation, stenting, or selected stone-related procedures. It is particularly relevant when endoscopic retrograde cholangiopancreatography is unsuccessful or anatomically difficult, although the choice of route varies with the clinical setting and local expertise. Obstructed infected bile can be associated with severe systemic illness, so the intervention’s purpose may be urgent decompression rather than immediate definitive treatment of the underlying lesion.
Percutaneous nephrostomy places a drainage catheter into the kidney’s collecting system. It can decompress an obstructed system, provide access for subsequent urinary procedures, or divert urine when normal drainage is disrupted. Causes of obstruction include kidney stones, tumors, and strictures. Antegrade ureteric stenting can sometimes restore internal drainage from the kidney to the bladder. Procedural imaging must account for the collecting system, adjacent bowel, pleura in upper-pole approaches, and the kidney’s vascular anatomy.
Feeding access, fluid management, and selected lymphatic interventions further illustrate the field’s breadth. Radiologically inserted gastrostomy establishes access to the stomach for enteral nutrition in appropriate circumstances. Lymphangiography can identify abnormal lymphatic channels, while embolization of the thoracic duct or other selected channels can treat some persistent lymphatic leaks. These procedures differ substantially in complexity and evidence base; their common feature is image-guided access to a target that would otherwise require a different route or operation.
Musculoskeletal interventions include bone biopsy, selected joint and soft-tissue procedures, ablation of painful lesions, and cement augmentation of some weakened bones. Vertebroplasty injects cement into a vertebral body; kyphoplasty includes cavity creation, commonly using a balloon, before cement delivery. Their effects in osteoporotic vertebral fractures have been examined in sham-controlled and other randomized trials, with findings influenced by patient selection, fracture timing, and trial design. Cement leakage, embolization, and injury to adjacent structures are recognized complications. These procedures therefore cannot be understood solely as mechanical reinforcement: the relationship between imaging abnormalities, fracture activity, and the source of pain is clinically important.
Periprocedural assessment and anesthesia
An interventional procedure begins with determining whether the proposed treatment addresses the clinical problem. The imaging abnormality, symptoms, physiological disturbance, and available alternatives may not align automatically. A narrowed artery, for example, is not by itself proof that opening it will improve organ function or survival. Assessment therefore includes the indication, the intended benefit, technical feasibility, competing treatments, and the consequences of doing nothing immediately. Consent addresses these issues together with the possibility of conversion to another treatment or failure to achieve the intended result.
Preprocedural evaluation draws on the medical history, examination, laboratory results, and prior imaging. Relevant factors include bleeding history, current antithrombotic drugs, renal function, infection, cardiopulmonary reserve, allergies, and previous surgery. Assessment of blood coagulation is procedure-specific: a superficial puncture and a deep solid-organ biopsy do not carry identical bleeding consequences. Decisions about interrupting antithrombotic therapy involve a balance between procedural bleeding and the risk of thrombosis. Laboratory thresholds alone do not capture every aspect of that balance.
Anesthesia ranges from local anesthetic infiltration to minimal or moderate sedation, deep sedation, and general anesthesia. The choice depends on expected discomfort, procedure duration, required immobility, airway considerations, and the patient’s condition. Thermal ablation, complex endovascular repair, and some pediatric procedures may require general anesthesia, while many access and drainage procedures can be performed under local anesthesia with or without sedation. Sedation is a continuum rather than a fixed state; respiratory depression or loss of airway control can occur even when deep anesthesia was not intended.
Procedural care requires monitoring and an environment capable of responding to deterioration. Depending on the procedure and anesthetic depth, monitoring can include blood pressure, electrocardiography, oxygen saturation, and exhaled carbon dioxide. Sterile technique, appropriate equipment preparation, correct-site verification, and explicit communication among operators, nurses, radiographers, and anesthesia staff are part of patient safety. Antibiotic prophylaxis is used for selected procedures according to their contamination risk and the consequences of infection, rather than uniformly for every intervention.
Postprocedural assessment is matched to the mechanism of possible harm. After arterial access, observation includes bleeding, limb perfusion, and the access site. After lung biopsy, assessment may focus on pneumothorax and pulmonary bleeding. Following biliary drainage, concerns include infection, bleeding, catheter function, and ongoing obstruction. Discharge from a procedural unit does not end clinical responsibility: drain exchanges, stent surveillance, tumor-response imaging, access maintenance, and recognition of delayed complications may require continuing specialist follow-up.
Radiation, contrast agents, and other procedural risks
Radiation exposure is a distinctive concern in fluoroscopic and CT-guided interventions. Tissue reactions, such as skin injury, occur above dose-dependent thresholds, whereas the probability of radiation-induced cancer is treated as increasing with dose without a sharp threshold in standard radiation-protection models. Complex procedures can concentrate radiation on a limited skin region, especially when similar projections are used repeatedly. Displayed reference air kerma and kerma–area product provide useful monitoring information but are not identical to peak skin dose or an individual patient’s effective dose.
Dose reduction combines equipment settings and procedural technique. Pulsed rather than continuous fluoroscopy, lower pulse rates when adequate, tight collimation, appropriate detector positioning, and avoidance of unnecessary image acquisitions can reduce exposure. Changing beam orientation may distribute skin dose, although its effects depend on geometry. Staff receive mainly scattered radiation; shielding, distance, and careful positioning reduce occupational exposure. Ultrasound and MRI avoid ionizing radiation, but substituting them is only useful when they provide adequate guidance for the intended task.
Iodinated contrast agents can produce allergic-like reactions ranging from mild skin manifestations to severe anaphylaxis. They also raise concerns in patients with impaired kidney function. An important distinction is between acute kidney injury occurring after contrast administration and injury causally attributable to contrast: acutely ill patients often have several simultaneous reasons for deteriorating renal function. Risk assessment depends on baseline kidney function, acute illness, contrast route and amount, and other clinical factors. For MRI-guided procedures, some gadolinium-based agents have been associated with nephrogenic systemic fibrosis in patients with severe renal impairment; risk differs substantially among agent groups.
Mechanical and biological complications are procedure-specific. Vascular access can cause hematoma, pseudoaneurysm, dissection, or thrombosis. Embolization can injure tissue outside the intended target if material enters the wrong branch or crosses an unexpected connection. Ablation may damage neighboring structures, while drainage and instrumentation can introduce infection or release infected material into the bloodstream. Postembolization syndrome—typically involving pain, fever, nausea, and inflammatory symptoms—must be distinguished from complications such as tissue infection or major organ injury.
Risk reporting is meaningful only when definitions and follow-up are explicit. “Technical success” may mean deployment of a device or completion of an intended embolization, whereas “clinical success” refers to the patient-centered objective, such as cessation of bleeding or resolution of obstruction. Complications can be classified by severity, treatment required, prolonged hospitalization, permanent harm, or death. Consistent reporting also includes late events, because restenosis, device failure, recurrent obstruction, and delayed infection may not be apparent during the initial admission.
Evidence, multidisciplinary practice, and service organization
The effectiveness of an intervention cannot be inferred solely from its technical elegance or small access incision. Randomized controlled trials can determine whether a procedure improves outcomes compared with medical treatment, surgery, another intervention, or, where ethically and practically appropriate, a sham procedure. Blinding is often difficult, and outcomes such as pain or functional improvement can be influenced by expectations and co-interventions. Trials must also distinguish improvements in a physiological measurement from reductions in disability, hospital admission, or death.
Renal artery stenting illustrates the difference between restoring anatomy and improving overall outcomes. The ASTRAL trial, published in 2009, and the CORAL trial, published in 2014, did not demonstrate broad clinical benefit from routine revascularization added to medical treatment in their studied populations with atherosclerotic renal artery disease. These results constrained general claims for the procedure, while leaving questions about less well-represented clinical presentations to more specific investigation. In contrast, multiple trials reported in 2015 established major benefits of endovascular thrombectomy for appropriately selected patients with acute ischemic stroke caused by large-vessel occlusion. Evidence therefore attaches to a defined indication, population, and time window, not to catheter treatment as a whole.
Device evaluation presents additional challenges. Devices can be modified more quickly than long-term comparative evidence accumulates, and operator experience may change outcomes during adoption. Registries and cohort studies can characterize uncommon complications, durability, and performance outside trial populations, but comparisons are vulnerable to selection bias and confounding. Patency, reintervention, symptom relief, organ preservation, and quality of life may each be important, yet they are not interchangeable. Reporting only the most favorable endpoint can obscure the practical burden of repeat procedures.
Clinical organization strongly affects whether an intervention can be delivered effectively. Emergency embolization requires timely imaging, appropriately trained personnel, suitable equipment, blood products when necessary, and access to surgical and intensive-care support. A thrombectomy service depends on stroke recognition, transport pathways, imaging selection, and coordinated emergency care, not simply an available angiography room. Elective interventional oncology similarly relies on pathology, oncology, surgery, imaging follow-up, and multidisciplinary review.
Training includes diagnostic imaging, radiation protection, vascular and nonvascular access, management of complications, and patient care. Competence is procedure-specific, and complex fields such as neurointervention or pediatric intervention require additional expertise. Facilities increasingly provide outpatient clinics, inpatient consultation, and structured follow-up rather than receiving patients only at the moment of a procedure. Access remains uneven across health systems, particularly where equipment maintenance, consumable supply, specialist staffing, and referral networks are limited. Service evaluation consequently includes not only procedural volume but also waiting times, emergency coverage, complication management, device retrieval, and continuity of care.