Choosing the right imaging study depends on the clinical scenario, the body part involved, and specific findings from the patient’s history and physical exam. There is no universal answer, but evidence-based decision rules and appropriateness criteria exist for nearly every common presentation. Understanding these frameworks prevents unnecessary scans, reduces radiation exposure, and gets patients to the right diagnosis faster.
How Clinicians Choose the Right Imaging Study
The American College of Radiology publishes Appropriateness Criteria that rate imaging options on a scale for hundreds of clinical scenarios. These physician-developed, evidence-based guidelines form the backbone of most clinical decision support tools now integrated into electronic health records. When a provider places an imaging order, the system cross-references the clinical indication against these criteria and flags studies that may be inappropriate.
This isn’t just advisory. The Protecting Access to Medicare Act of 2014 requires providers to consult a qualified clinical decision support tool when ordering advanced imaging (CT, MRI, nuclear medicine, PET) for Medicare patients. The system returns an appropriateness score, and orders rated as inappropriate may require prior authorization before proceeding. For orders rated as appropriate, the process often bypasses pre-authorization entirely, which can otherwise delay care by five to ten days.
Head Trauma: When CT Is Needed
Not every bump on the head requires a CT scan. Two validated clinical decision rules help determine which patients with minor head injuries actually need imaging.
The Canadian CT Head Rule flags patients as high risk if they have a reduced level of consciousness two hours after injury, a suspected open or depressed skull fracture, signs of a basal skull fracture (like bruising behind the ears or blood behind the eardrums), two or more episodes of vomiting, or are 65 years or older. Medium-risk features include memory loss of 30 minutes or more before the injury and a dangerous mechanism such as a car crash or a fall from height. Any of these findings warrants a CT scan.
The New Orleans Criteria cast a wider net: imaging is indicated if the patient has any headache, vomiting, intoxication, memory loss, a seizure, visible trauma above the collarbone, or is older than 60. Both rules have high sensitivity for detecting significant injuries, meaning they rarely miss a bleed. The Canadian rule is more specific, resulting in fewer scans overall.
Children Under 16
Pediatric head injuries follow a separate algorithm called the PECARN rule, which is divided by age. For children under two, a CT scan can be safely avoided if the child has a normal mental status, is behaving normally according to the caregiver, had no loss of consciousness, has no scalp swelling outside the forehead, no evidence of skull fracture, and no dangerous injury mechanism. For children aged 2 to 15, the criteria shift slightly: normal mental status, no loss of consciousness, no vomiting, no severe headache, no signs of a basilar skull fracture, and no dangerous mechanism. Meeting all criteria in the appropriate age group means the risk of a significant brain injury is very low, and observation is preferred over scanning to avoid unnecessary radiation in a developing brain.
Suspected Stroke: Speed Over Perfection
When stroke is suspected, a non-contrast CT of the head is the first scan ordered, and it should happen as soon as the patient is stabilized in the emergency room. The primary goal at this stage isn’t to see the stroke itself. It’s to rule out bleeding, because the treatment for a clot (blood thinners or clot-retrieval procedures) would be catastrophic if the problem is actually a hemorrhage.
Non-contrast CT is fast, widely available, and excellent at detecting bleeding. It is, however, limited in showing early ischemic strokes. That’s where the next steps come in. Many stroke centers follow with CT angiography to locate a blocked vessel and CT perfusion to map which brain tissue is still salvageable. This combination of non-contrast CT, CT angiography, and CT perfusion forms the standard “code stroke” imaging protocol.
MRI with diffusion-weighted imaging is actually the most sensitive and specific tool for detecting ischemic stroke, picking up infarction within minutes of onset with sensitivity between 88% and 100% and specificity between 95% and 100%. But MRI takes longer, isn’t available around the clock at every facility, and in the hyperacute phase, speed matters more than image quality. MRI becomes more valuable in the subacute phase, when confirming the diagnosis and assessing the full extent of injury guides ongoing treatment decisions.
Pulmonary Embolism: A Stepwise Approach
Suspected blood clots in the lungs follow one of the most structured imaging decision pathways in medicine. The process starts not with imaging but with risk stratification using a scoring system like the Wells Score or the Revised Geneva Score, which estimate the probability of a pulmonary embolism based on clinical features like heart rate, recent surgery, leg swelling, and whether an alternative diagnosis is more likely.
For patients who score as low risk, the PERC rule can be applied first. This is a set of eight criteria: age under 50, heart rate under 100, oxygen saturation 95% or above, no leg swelling, no coughing up blood, no recent surgery or trauma within four weeks, no estrogen use, and no prior blood clots. If a low-risk patient meets all eight, pulmonary embolism can be excluded on clinical grounds alone, with no blood tests or imaging needed.
If any PERC criterion is not met, a D-dimer blood test is the next step. A negative high-sensitivity D-dimer safely rules out pulmonary embolism in low and intermediate-risk patients. A positive D-dimer, however, triggers imaging, typically CT pulmonary angiography. For patients who score as high risk on initial assessment, imaging should be ordered immediately without waiting for D-dimer results, since the blood test isn’t reliable enough to rule out a clot when clinical suspicion is high.
Abdominal Pain: Location Determines Modality
The right imaging study for abdominal pain depends heavily on where the pain is and what’s suspected. For right upper quadrant pain, ultrasound is the primary imaging modality. It is both sensitive and specific for detecting gallstones, dilated bile ducts, and signs of acute inflammation like gallbladder wall thickening or fluid around the gallbladder. CT becomes useful as a follow-up if complications are suspected, such as perforation, abscess, or involvement of surrounding structures, but ultrasound comes first.
For left lower quadrant pain in adults, where diverticulitis is a common concern, CT with contrast is generally the first-line study. CT provides excellent visualization of the colon, surrounding fat, and any abscess or perforation. For right lower quadrant pain suggesting appendicitis, CT is also the standard in adults, though ultrasound is preferred as the initial study in children and pregnant patients to minimize radiation exposure.
Low Back Pain: Less Imaging Is Usually Better
Imaging for low back pain is one of the most frequently overused tests in medicine. Only 5% to 10% of low back pain presentations in primary care involve a condition that actually requires imaging. For the vast majority of patients, scans don’t improve outcomes. Research shows that people who receive imaging for low back pain incur higher medical costs, use more healthcare services, and miss more work compared to those who don’t get scanned.
International “Choosing Wisely” campaigns, the Canadian Spine Society, the Australian Rheumatology Association, and the UK’s Royal College of Physicians all agree: do not routinely image patients with low back pain regardless of how long it has lasted, unless red flags are present. Plain X-rays of the lumbar spine expose patients to ionizing radiation and provide no benefit in terms of pain, function, or disability for non-specific back pain. MRI is similarly not recommended for uncomplicated cases.
The red flags that do justify imaging include suspected cancer (unexplained weight loss, history of malignancy), infection (fever, IV drug use, recent procedure), inflammatory disease, fracture (significant trauma, osteoporosis, steroid use), and severe or progressive neurological deficits like loss of bowel or bladder control or rapidly worsening leg weakness. These warrant prompt imaging, typically MRI, to identify conditions that need specific treatment.
Ankle and Foot Injuries: The Ottawa Rules
The Ottawa Ankle Rules are among the most validated clinical decision rules in emergency medicine, designed to determine whether an X-ray is necessary after an ankle or foot injury. An ankle X-ray is indicated if the patient is 55 or older, cannot bear weight for four steps both immediately after the injury and at the time of the exam, or has bone tenderness along the back edge or tip of either the inner or outer ankle bone. A foot X-ray is indicated if there’s bone tenderness at the base of the fifth metatarsal (the outer edge of the midfoot), the cuboid, or the navicular bone. If none of these criteria are present, the chance of a fracture is extremely low and X-rays can be safely skipped.
Imaging During Pregnancy
Ultrasound and MRI are the preferred imaging techniques for pregnant patients because neither involves ionizing radiation. However, the American College of Obstetricians and Gynecologists is clear that X-rays, CT scans, and nuclear medicine studies should not be withheld from a pregnant patient when they are medically necessary. With few exceptions, the radiation doses from diagnostic imaging fall well below levels associated with fetal harm. For context, a fetus is exposed to roughly 1 milligray of background radiation over the course of a normal pregnancy, and most diagnostic studies deliver doses in that same range or only modestly above it.
When MRI is used in pregnancy, gadolinium-based contrast agents are generally avoided unless the information is critical and unobtainable any other way. The FDA requires warnings on all gadolinium-based contrast agents noting that gadolinium can remain in body tissues, including the brain, for months to years after injection. In patients with normal kidney function, this retention has not been directly linked to adverse health effects, and the benefits of contrast-enhanced MRI continue to outweigh the risks. In patients with severe kidney impairment, however, gadolinium carries a risk of nephrogenic systemic fibrosis, a rare but serious condition. Kidney function is checked before administering gadolinium, and patients with significantly reduced filtration rates require alternative approaches or additional protective measures.
Contrast Safety and Kidney Function
For CT scans requiring iodinated contrast (the dye used to light up blood vessels and organs), kidney function matters. The risk of contrast-related kidney injury rises meaningfully when the kidneys are already filtering below about 60 mL/min, and it becomes a serious concern below 30 mL/min, especially in patients who also have diabetes. Patients at intermediate risk typically receive IV fluids before and after the scan to protect the kidneys, while those at highest risk may need nephrology input and consideration of alternative imaging that doesn’t require contrast.
The same principle of using the lowest effective dose applies to radiation. Choosing ultrasound or MRI when they can answer the clinical question avoids radiation entirely. When CT is necessary, modern scanners use dose-reduction techniques, and the clinical benefit of making the right diagnosis almost always outweighs the small radiation risk from a single study.