CNMT common mistakes
CNMT · Understand the traps and how to avoid them.
Calculating decay-corrected activity by multiplying the current activity by 0.5 raised to (elapsed time/half-life) when the question asks for the activity that must have been present at an earlier time, producing a value below the current reading instead of above it.
The exponential factor is applied mechanically without checking whether the clock is running forward or backward; decay-corrected dose preparation is practiced in one direction so the inverse case feels unfamiliar.
What to do instead
Before computing, write down the direction: if the question moves forward in time, activity decreases (multiply by the decay factor); if it moves backward, activity increases (divide by the decay factor). Sanity-check the answer against the starting value before selecting an option.
Radiopharmaceutical quality-control questions answered from memory of a single number (e.g., 90% for Tc-99m radiochemical purity) without distinguishing radiochemical, radionuclidic, and chemical purity or the specific assay method named in the stem.
QC limits are memorized as isolated facts during review rather than as a matrix of impurity type, test method, and acceptance limit, so any stem that varies one axis breaks recall.
What to do instead
Build a one-page grid with rows for radiochemical, radionuclidic, and chemical purity and columns for the test method (chromatography vs. dose-calibrator/multichannel analysis), the impurity measured, and the acceptance limit; recite the grid until each cell can be produced independently.
Selecting a patient preparation or pregnancy/breastfeeding instruction for a nuclear medicine procedure that contradicts the actual administered agent, such as advising a specific breastfeeding interruption interval that does not match the radiopharmaceutical in the stem.
Preparation instructions are learned as generic lists ('hold this, hydrate, void') rather than tied to the physical characteristics of each tracer, and exam stems deliberately pair a familiar instruction with a different agent.
What to do instead
For each commonly tested tracer, write the agent name, its effective half-life and route of excretion, and the corresponding patient instruction on the same line; review by covering the instruction column and regenerating it from the agent.
Applying a single CT window/level or reconstruction kernel to every protocol instead of matching window width, window level, slice thickness, and algorithm to the clinical task (e.g., lung vs. bone vs. soft tissue vs. attenuation-correction maps).
Candidates memorize one 'good' display setting from a workstation demonstration and reuse it, missing that the post-primary CT content tests task-based parameter selection rather than a default.
What to do instead
Practice by stating, for each anatomy and indication, the window width, window level, slice thickness, and kernel, and explain in one sentence why each value changes; drill until the rationale, not the number, drives the answer.
Misidentifying CT artifacts by appearance and cause, for example calling beam-hardening streaks 'motion' or attributing ring artifacts to patient motion, then choosing the wrong correction.
Artifacts are studied as pictures with labels rather than as physical mechanisms, so when a stem describes the appearance in words the candidate cannot map description to cause.
What to do instead
For each artifact, memorize the triplet: physical cause, characteristic appearance in words, and the practical mitigation (e.g., higher kVp or iterative correction for beam hardening; patient coaching or faster rotation for motion; detector calibration for rings). Quiz yourself from the cause side and from the appearance side.
Reporting CT dose metrics interchangeably, quoting CTDIvol where the stem asks for a size-adjusted or whole-scan dose descriptor, or failing to account for pitch and scan length when estimating dose.
CTDIvol, dose-length product, and size-specific dose estimates are learned as a cluster of acronyms without the relationship among them, and pitch is treated as an image-quality parameter only.
What to do instead
Write the relationships explicitly (dose-length product equals CTDIvol times scan length; pitch affects CTDIvol for a given protocol) and work several numeric problems where one input changes at a time so the effect of each variable is isolated.
Interpreting a myocardial perfusion study without integrating the stress modality used, the patient's medications and caffeine status, and the presence of attenuation or motion artifact before declaring a defect reversible or fixed.
Candidates jump from image pattern to diagnosis, skipping the acquisition and preparation context that determines whether a defect is real, and the exam rewards systematic elimination of artifacts first.
What to do instead
Adopt a fixed reading sequence for every case: verify preparation and stress adequacy, check for attenuation, motion, and extra-cardiac uptake, then assess defect location, severity, and reversibility, and only then commit to an interpretation. Practice stating the sequence aloud on each practice case.
Confusing the acquisition and processing requirements of gated SPECT versus first-pass or planar equilibrium studies, such as expecting ejection fraction to come from the same reconstruction pipeline as perfusion images.
Functional and perfusion cardiac imaging are reviewed in separate blocks, so the boundary between them (what each study measures, how it is acquired, what can go wrong) is never tested as a comparison.
What to do instead
Create a comparison table with columns for acquisition mode, radiopharmaceutical and bolus requirements, processing steps, and derived parameters for each cardiac study type; fill it from blank repeatedly until the distinctions are automatic.
Reporting SUV as a stand-alone diagnostic number without reference to lesion size, blood-pool or liver background, time from injection, and scanner/processing conditions.
SUV is treated as a fixed threshold (a single cutoff memorized for malignancy) rather than a semi-quantitative index whose meaning depends on acquisition and analysis context.
What to do instead
For every SUV question, explicitly state the reference region, the time point, and the confounders (small lesion size, respiratory motion, high background, diabetes/insulin status) before interpreting; practice by writing the caveats for a given SUV value rather than the value alone.
Applying a single fixed blood-glucose cutoff and fasting interval to every FDG indication without adjusting for the specific clinical question, the patient's diabetes management, and the possibility of a nondiagnostic but not necessarily canceled study.
Preparation rules are memorized as one universal number, and the distinction between 'reschedule', 'proceed with caveats', and 'cancel' is not practiced.
What to do instead
Build a decision tree that branches on indication, glucose value, and diabetes regimen, and rehearse the wording of the resulting recommendation for each branch so the answer reflects the clinical question rather than a single threshold.
Accepting a radiopharmaceutical dose or patient identity without performing the independent verification steps expected of an advanced associate, such as confirming the prescribed activity against the dose calibrator reading and the requisition at the point of administration.
Assistants often learn workflow as a sequence of tasks to complete quickly and treat verification as a formality already done upstream, so questions framed around catching an error are answered with the 'normal' step instead of the check.
What to do instead
Rehearse the administration pathway as a series of independent confirmations (identity, prescribed vs. measured activity, agent, route, time), and for each step name what could go wrong if it were skipped; answer exam items by identifying the next verification that would detect the discrepancy.
Handling a radiopharmaceutical spill or a contaminated package by improvising a response instead of following the ordered steps of containment, survey, decontamination, and documentation, and misjudging when to notify the radiation safety officer.
Emergency procedures are read once as prose and never converted into an ordered checklist, so under exam pressure the candidate selects a plausible-sounding action that skips a step or reverses the order.
What to do instead
Write the spill and package-receipt response as numbered steps from memory, including survey instrument use, area restriction, and reporting thresholds, then re-derive the list weekly until the sequence is reliable.