ABPS Diagnostic Radiology common mistakes
ABPS Diagnostic Radiology · Understand the traps and how to avoid them.
Committing to a single diagnosis for a cavitary or cystic lung lesion without first generating the full differential (infection, malignancy, vasculitis, congenital), then answering the follow-up question about the next management step based on that single anchor.
Candidates recognize a classic pattern (e.g., upper-lobe cavitation) and stop reasoning; the exam then tests the discrimination between mimics, and the anchored answer propagates through linked questions.
What to do instead
Before selecting, explicitly list at least three diagnostic categories for any cavitary/cystic lesion and identify the single discriminating feature (wall thickness, dependent layering, multiplicity, associated adenopathy) that separates them.
Describing a bowel obstruction pattern without systematically localizing the transition point and stating whether a closed loop or strangulation is present.
Trainees default to 'dilated loops = SBO' and skip the decompressed-distal-bowel and loop-configuration analysis that the question actually hinges on.
What to do instead
For every obstruction case, force the sequence: dilated vs decompressed segments, transition zone location, loop shape (U/C vs radial), wall enhancement, and mesenteric vessels-then commit to open vs closed loop before answering.
Assigning a renal mass to a histologic subtype based on a single feature (e.g., macroscopic fat) without checking whether the feature is truly present versus chemical-shift artifact or lipid-poor composition.
Candidates memorize 'fat = angiomyolipoma' and do not verify true fat (signal drop on fat-sat, India ink artifact at interfaces) versus intracellular lipid, so they misclassify lipid-poor RCC or fat-containing mimics.
What to do instead
For each renal lesion, document enhancement (pre/post), true macroscopic fat confirmation on fat-suppressed sequences, and T2 signal; answer only after the subtype features are individually confirmed.
Reading a bone or joint lesion without first deciding benign versus aggressive using margin, zone of transition, periosteal reaction, and matrix, instead of jumping to a named entity.
Pattern-recognition pressure under timed conditions leads to naming a diagnosis from a single salient feature (e.g., 'sunburst = osteosarcoma') and ignoring contradictory features.
What to do instead
Apply a fixed checklist (margin/zone of transition, cortical destruction, periosteal type, matrix, soft-tissue component, patient age) and state benign vs aggressive before naming any entity.
Localizing a stroke or lesion on cross-sectional brain imaging by gestalt rather than tracing vascular territories and identifying the specific artery and supplied structures.
Candidates recognize 'MCA territory' broadly but cannot distinguish perforator, cortical, or watershed patterns, which is exactly what the discriminating question tests.
What to do instead
For each infarct or lesion, name the vascular territory, the specific artery, and the supplied anatomic structures involved; verify the pattern against known watershed and perforator distributions before answering.
Calling a breast finding BI-RADS 3 (probably benign) when the morphology or stability criteria for short-interval follow-up are not actually met, or conversely escalating a definitively benign finding.
Candidates apply BI-RADS categories by familiarity rather than by the explicit morphologic and stability criteria, and they conflate 'looks benign' with 'meets follow-up criteria'.
What to do instead
For every breast finding, state the specific descriptor (shape, margin, density, calcification morphology/distribution) and match it to the exact BI-RADS criterion; only assign 3 when the defined probably-benign criteria are satisfied.
Interpreting a cardiac or vascular study without measuring the quantitative parameter the question depends on (e.g., aortic diameter indexed to body size, valve area, ejection fraction, or stenosis severity by the correct criterion).
Candidates rely on visual impression and skip the measurement or the guideline threshold, so they misjudge severity categories that determine management.
What to do instead
Before answering, identify the parameter being tested, recall the guideline threshold (e.g., aortic aneurysm size thresholds, stenosis grading criteria), and state the measured value against that threshold.
Applying adult imaging patterns to pediatric cases without adjusting for age-specific normal variants (e.g., thymus, physiologic periosteal reaction, open physes, normal bowel gas patterns).
Most training exposure is adult; candidates misread normal pediatric appearances as pathology or miss age-specific disease presentations.
What to do instead
For every pediatric image, first state the patient's age and the expected age-specific normal findings, then evaluate for pathology relative to that baseline.
Interpreting a nuclear medicine or PET study without confirming the radiopharmaceutical, its biodistribution, and the correct semiquantitative or visual criterion being applied.
Candidates treat all uptake as comparable and ignore tracer-specific normal distribution and physiologic variants, leading to false-positive or false-negative calls.
What to do instead
For each NM/PET question, identify the tracer, state its normal biodistribution and known pitfalls, then apply the study-specific interpretation criterion (e.g., visual grading, SUV threshold, or comparison to background) explicitly.
Answering an obstetric or gynecologic ultrasound question without correlating the finding to gestational age, dating criteria, and the correct discriminatory measurement (e.g., CRL vs nuchal translucency vs cervical length).
Candidates recognize a structure but not the gestational-age-dependent normative threshold, so they misclassify normal versus abnormal at a given age.
What to do instead
For each OB/GYN question, state gestational age, the relevant measurement, and its age-specific normal range or cutoff before choosing an answer.
Selecting an interventional or vascular procedure or device without verifying the indication, contraindication, and anatomic suitability the question is testing (e.g., access site, vessel caliber, coagulation status).
Candidates match a disease to a procedure by association and skip the safety and suitability checks that distinguish the correct from the plausible-but-wrong option.
What to do instead
For each IR/vascular question, list the indication, absolute contraindications, and anatomic/physiologic prerequisites, and eliminate options that violate any of them before selecting.
Computing or reasoning about radiation dose, shielding, or image quality without setting up the governing physical relationship (inverse square, half-value layer, mA·s, or dose-length product) before choosing an answer.
Candidates answer physics items from memory of a formula's name rather than applying the actual relationship to the numbers given, so unit and proportionality errors go undetected.
What to do instead
For every physics/QA calculation, write the governing equation and units, substitute the given values, and sanity-check the direction and magnitude of the result before selecting.