Attenuation & Misregistration Artifacts
When the correction, the CT, or the patient, not the tracer, creates the finding
Quick Answer
Many false findings on hybrid imaging come from the attenuation-correction (AC) process and from misregistration between the emission and CT data, not from tracer distribution. Dense material (metal, iodinated contrast, calcium), breathing motion, and patient positioning can create apparent uptake where there is none or erase real uptake. The universal defense is to compare the attenuation-corrected and non-attenuation-corrected (NAC) images and inspect the fused CT: an artifact behaves differently across those datasets than true disease.
Why AC can lie
PET and SPECT reconstruct quantitative images by using the CT (or a transmission map) to correct for photon attenuation. That correction assumes the emission and attenuation data describe the same anatomy in the same place. When they do not — because the patient breathed, moved, or has dense material the CT over-weights — the algorithm redistributes counts incorrectly, producing spurious hot or cold regions. The corrected image is only as trustworthy as the alignment and the CT it was built from.
The core artifacts
Misregistration (respiratory mismatch). The classic example is a curvilinear cold or hot band at the lung base/dome of the diaphragm/liver, where CT (a fast breath-hold snapshot) and PET (averaged over many breaths) capture the diaphragm in different positions. A hepatic-dome lesion can be shifted into lung or vice versa. Checking the NAC images and the fused display — where the mislocation is obvious — resolves it.
Dense-material (CT contrast, metal, calcium) over-correction. High-attenuation material makes the CT assign excessive attenuation to that voxel, so AC over-corrects and creates apparent uptake on the AC image that is absent on NAC. Culprits: IV/oral CT contrast, hip prostheses and dental hardware, pacemakers, chemo ports, and contrast in the renal collecting system. The tell is focal "uptake" that vanishes on the NAC series and co-locates with dense material on CT.
Halo artifact. A rim of apparent decreased activity around an intensely avid focus (e.g. a brightly avid tumour or the bladder), an over-/under-correction reconstruction effect that can mimic or obscure adjacent disease.
Truncation artifact. When the patient's body extends beyond the CT field of view (arms at the sides, large body habitus), the missing anatomy is under-corrected, producing streaks or falsely altered peripheral activity.
Cardiac SPECT MPI attenuation
On myocardial perfusion SPECT, soft-tissue attenuation creates fixed apparent perfusion defects that mimic infarct: anterior/anterolateral (breast) attenuation in women and inferior/inferolateral (diaphragm) attenuation in men. Two tools separate artifact from scar: gated wall-motion (an attenuation artifact has normal wall motion and thickening; true scar is hypokinetic/akinetic) and attenuation correction or prone/upright imaging that redistributes the soft tissue. A fixed defect with preserved motion is the signature of attenuation, not infarction.
Other MPI pitfalls in the same family: patient motion (blurring, "hurricane" sign), subdiaphragmatic activity (bowel/liver) creating reconstruction ramp artifacts adjacent to the inferior wall, and breast-position variation between stress and rest that mimics a reversible defect. Recognising these prevents both false-positive catheterisations and missed disease.
The universal workflow
- Always review NAC alongside AC images. Artifact-driven "uptake" from dense material disappears on NAC; true disease persists on both.
- Inspect the fused CT for metal, contrast, and diaphragm alignment.
- For cardiac SPECT, read the gated cine and use AC/prone imaging before calling a fixed inferior or anterior defect.
- When registration is poor, consider re-processing or repeating rather than over-interpreting.
Self-Check
Q1. Why can attenuation correction create a false hot spot from CT contrast or metal?
Answer: Dense material makes the CT assign excessive attenuation to that voxel, so AC over-corrects and adds apparent counts. The spurious "uptake" is present on the AC image but absent on the non-attenuation-corrected (NAC) series and co-locates with the dense material on CT.
Q2. What causes the curvilinear defect at the lung base/diaphragm, and how is it recognised?
Answer: Respiratory misregistration — CT (breath-hold) and PET (free-breathing average) capture the diaphragm in different positions, mislocating activity between liver dome and lung base. The fused display and NAC images reveal the mismatch.
Q3. On cardiac SPECT, how do you distinguish breast/diaphragm attenuation from true infarct?
Answer: A fixed defect with preserved wall motion and thickening on the gated images (plus resolution with attenuation correction or prone/upright imaging) indicates attenuation; true scar shows abnormal (hypo/akinetic) motion.
Q4. What single habit defends against most attenuation/misregistration artifacts?
Answer: Reviewing the non-attenuation-corrected images alongside the AC images and inspecting the fused CT — artifacts behave differently across those datasets than true tracer uptake.
Key References
- SNMMI/EANM guidance on PET/CT and SPECT/CT attenuation correction, misregistration, and artifact recognition.
- Nuclear cardiology literature on soft-tissue attenuation, gated wall-motion analysis, and attenuation-correction/prone imaging for MPI.
Evidence & sources
https://nuc.med/attenuation-artifacts/ Report an issue