Artifacts--Pearls Flashcards
PM tube malfunction
Cold spot in floods
Ring artifact in ? which images
Streak artifacts in final reconstructed tomographic short-axis images
Usually fixed defects, though may appear reversible if LV is positioned differently in field of view, so exact location of PMT vis a vis LV is somewhat different
K
COR error = same appearance as motion artifact
1) oblong LV cavity
2) opposed defects (AW and IW if vertical motion; septum and LW if horizontal motion)
3) streaks of activity coming off the heart (“hurricane”)
You should NOT interpret motion images; either see if completely corrected by motion correction software, or repeat study
Lateral (horizontal) motion:
may be difficult to see on projection images;
Can see it clearly on sinogram as a disconnect in the border (sinogram is stacked images of septum and lateral wall)
K
Errors in processing raw images
Error in selection of apex/base limits:
Not just causes defects in polar plots but also affects SSS/SDS calculation, LV volume calculation, EF, TID ratio
Error in LV axis alignment:
If bottom row images (HLA) are not pointed straight up (to 12:00), it means axis is off and may get artifactual defects
K
Ramp filter artifact
Clues it’s ramp filter and not true inferior ischemia:
1) see good counts in inferior wall on raw projection spinning images
2) see the strong subdiaphragmatic tracer activity
3) ? large black spot on tomographic images
If in liver–remedy is just wait long enough (2 hrs?)
If stomach or bowel–can drink water and ambulate to speed things along
K
Attenuation Artifacts
Typically WORSE on REST images, bc accentuated in low count density images
Looks like reverse redistribution but it’s pseudo-
Real reverse redistribution would have abnormal wall motion/thickening
(occurs when subendocardial infarct, i.e. not full thickness, with patent subtending vessel…so with hyperemia, the signif increased counts to remainder of the LV wall obscures the contrast btw the abnormal segment and neighboring segments)
If large, low-hanging breasts, the attenuation can be over inferior wall if that’s where densest breast tissue overlies (raw projection images will clue you in)
Very large, dense breasts:
See streaks of activity extending from tip of apex
Clue is big, dense black spot top right on bottom row images, from the L breast shadow
Diaphragmatic attenuation:
1) can’t see inferior wall clearly all the way to basal portion on raw projection image
2) may see subdiaphragmatic activity overlying IW on raw projection images
K
Variable breast attenuation artifact
Clues:
1) see the differences in where breast border is on raw images–higher up, intersecting AW on stress images, lower down covering more of, if not all of, LV on rest images
2) may look like reverse redistribution of apex bc denser breast tissue overlying apex on rest when hanging lower
No way to tell if variable breast vs LAD stenosis so must call it “indeterminate study”.
Should reimage pt and reposition their arms
K
LBBB artifact
Septum actually getting less perfusion due to its contracting during part of diastole (when CBF occurring)
Accentuated at higher HR so worse with exercise or dobutamine–reversible septal defect
K
Short septum anatomical variant
If looks like reversible BASAL septal defect (esp in short axis slices), this is not physiologic (bc has dual blood supply)…think short septum.
Look at HLA image slices (verticals) to confirm
Should be fixed defect, but if stress is half a slice off from rest it may look reversible
K
LVH
Septum thickest even in concentric LVH, so most counts–get hot spot
Remaining LV can look diffusely ischemic
K
RV insertion artifact
Anteroseptal (11 o’clock), but often also inferoseptal (7 o’clock), where RV inferior wall inserts
Often the defect is less apparent on rest images (bc use lower cut-off frequency for lower-count rest images, which blurs the rest images more)–looks partially reversible
? how can you differentiate it from true anteroseptal ischemia…is it much more discrete defect than true LAD stenosis, which would involve more of septum (eg down to 9 o’clock)…I think so
K
