Iran Specialty Board Radiology common mistakes
Iran Specialty Board Radiology · Understand the traps and how to avoid them.
Memorizing the 0.5-Tesla threshold for the MRI 'missile effect' and the CTDI-vol vs DLP formulas as isolated numbers, then applying them to the wrong modality or mixing CTDI-vol (mGy) with DLP (mGy·cm) in dose calculations.
Physics and radiation-protection facts are taught as a list of constants, and candidates rehearse them without connecting each quantity to its unit, its measurement geometry, or the exam stem's actual question (patient dose vs scanner output).
What to do instead
Build a one-page dimensional map: CTDI-vol = scanner output per rotation (mGy), DLP = CTDI-vol × scan length (mGy·cm), effective dose = DLP × k-factor (mSv). For each constant you memorize, write the unit and one clinical consequence next to it, then re-derive the formula from units when a stem is ambiguous.
Reading a chest CT or radiograph as a search for one 'dominant' finding and stopping when it is found, missing a second actionable lesion (e.g., a small pneumothorax, a rib lesion, a contralateral nodule) that the question is actually testing.
Radiology teaching rewards 'the diagnosis'; candidates carry a satisfaction-of-search shortcut from case-based study and under-train systematic review of every compartment, especially when the stem gives a compelling clinical vignette.
What to do instead
Adopt a fixed anatomical checklist (apices, pleura, mediastinal contours, hila, airways, parenchyma by zone, bones, soft tissues, tubes/lines) and verbalize it on every practice case. Deliberately practice 'second-finding' cases where the obvious lesion is benign and the answer is the incidental one.
Treating a widened mediastinum or a high-attenuation crescent on non-contrast CT as diagnostic of aortic dissection without checking the contrast phase, the presence of a true false-lumen flap, or the alternative diagnoses (intramural hematoma, penetrating ulcer, artifact from motion).
Pattern recognition from classic images creates an anchoring bias; candidates do not systematically verify the technical adequacy of the study before committing to the diagnosis.
What to do instead
For every acute-aorta stem, run a three-step check: (1) is contrast timing adequate for the structure in question, (2) is there a flap/intimal defect or only wall thickening, (3) what else could produce this (artifact, hematoma, ulcer, aneurysm with thrombus). Practice distinguishing dissection vs IMH vs PAU in the same case set.
Answering biliary and pancreatic questions by naming the most famous association (e.g., 'double-duct sign = pancreatic head cancer') without weighing the demographic and laboratory data in the stem that point to a different cause (stone, chronic pancreatitis, choledochal cyst, autoimmune pancreatitis).
High-yield association lists are over-learned, and candidates skip the discriminating data (age, CA 19-9, amylase, prior surgery) because they recognize the buzzword and move on.
What to do instead
For each hepatobiliary vignette, explicitly write the two or three findings that argue against the 'famous' diagnosis before selecting an answer. Build a differential table for each sign (double duct, biliary dilatation without stone, pancreatic mass) with the single best discriminator for each entry.
Confusing the imaging appearances of common renal and adrenal entities by relying on a single feature (e.g., calling any lipid-poor adrenal mass an adenoma, or any cystic renal lesion with septa a Bosniak III) instead of applying the full criteria set.
Candidates memorize one 'rule-out' feature (lipid content, septation) and do not rehearse the complete Bosniak or adrenal washout algorithm, including the thresholds for absolute and relative washout.
What to do instead
Reconstruct the Bosniak and adrenal-washout algorithms from first principles on paper: list every criterion, the cutoff values (e.g., >60% absolute washout at 15 min), and one example image description per category. Then answer stems by scoring the lesion against the checklist rather than pattern-matching.
Interpreting a musculoskeletal radiograph or MRI as a single-lesion exercise without localizing the lesion to a compartment and without using age and skeletal maturity to narrow the differential.
Trainees jump to a named diagnosis ('osteosarcoma') from the lesion's appearance and ignore the systematic approach (bone vs soft tissue, epiphysis/metaphysis/diaphysis, age, margin, periosteal reaction) that the exam expects to see applied.
What to do instead
Practice a written template for every bone lesion: age, location (which bone, which segment), margin (narrow/wide zone of transition), matrix, periosteal reaction, soft-tissue component. Answer choices by elimination against the template, and review pediatric vs adult differentials side by side.
On neuro cases, describing the finding but failing to commit to the anatomical localization (e.g., distinguishing a lesion in the basal ganglia from one in the thalamus or insula) and therefore choosing a diagnosis that does not fit the tract or vascular territory.
Candidates under-use cross-sectional anatomy in the exam setting; they recognize the pathology but not the precise structure, and stems often hinge on territory-based reasoning (e.g., MCA vs ACA vs PCA infarct).
What to do instead
Drill vascular territories and eloquent structures on normal scans until you can name them without a reference. For each stroke or mass stem, state the territory/structure in one sentence before looking at the options, then eliminate any option that does not respect that localization.
Applying BI-RADS categories inconsistently - for example, calling a circumscribed oval mass with a thin capsule 'BI-RADS 4' or assigning BI-RADS 3 to a lesion that has already changed - instead of using the lexicon descriptors and the management table.
Candidates learn BI-RADS as a list of categories without the descriptor-to-category logic, and they confuse 'probably benign' with 'needs biopsy' under exam time pressure.
What to do instead
Memorize the descriptor-to-category decision table (shape, margin, density, associated features) and the management linked to each category. Practice 20 stems where the same lesion is described with one descriptor changed, and predict how the category and management shift.
Answering pediatric radiology questions with adult thresholds and adult disease frequencies - e.g., treating a pediatric head ultrasound finding, a normal thymus, or a toddler's fracture with adult criteria - and missing age-specific normal variants.
Most study material and clinical exposure is adult; candidates under-weight the age in the stem and do not have a mental library of pediatric normal variants and age-specific injury patterns.
What to do instead
Maintain a dedicated pediatric section: normal variants by age (thymus, physiologic periosteal reaction, growth plates), age-specific fracture patterns (toddler's, supracondylar, non-accidental trauma), and pediatric-specific dose concerns. Before answering any pediatric stem, state the patient's age and what it changes.
Selecting an interventional or nuclear medicine answer on the basis of the procedure's name or the radiopharmaceutical's popularity without checking the indication, contraindication, or the specific uptake pattern that the stem describes.
Vascular/interventional and nuclear medicine content is often learned as procedure lists and tracer lists; candidates do not rehearse the decision rules (e.g., when to embolize vs observe, which tracer for which tumor, HIDA vs cholescintigraphy timing).
What to do instead
For each procedure and tracer, write an indication–contraindication–expected-finding triad and one clinical scenario where it is the wrong choice. For scintigraphy stems, explicitly match the described biodistribution and timing to the tracer before selecting the diagnosis.