Rc. Weichselbaum et al., Relevance of sonographic artifacts observed during in vitro characterization of urocystolith mineral composition, VET RAD ULT, 41(5), 2000, pp. 438-446
Nine pure mineral type canine uroliths (bladder or urethral origin only) we
re imaged ultrasonographically using 3.5 MHz, 5.0 MHz , and 7.5 MHz fixed f
ocus, mechanical sector transducers in a urinary bladder phantom, The uroli
ths studied were those composed of 100% magnesium ammonium phosphate, calci
um oxalate monohydrate, calcium oxalate dihydrate, calcium phosphate appati
te, and calcium hydrogen phosphate dihydrate (brushite), ammonium acid urat
e, sodium acid mate, cystine, and silica. The occurrence of both reverberat
ion/streak and acoustic shadowing artifacts were compared to urocystolith m
ineral type (classified by effective atomic number), urocystolith width, ur
ocystolith height (thickness), and ultrasonographic imaging frequency. No p
redictable relationship was found between either of the artifacts seen beyo
nd the urocystolith (reverberation/streak or the acoustic shadowing) and ur
ocystolith mineral type. There was no statistical relationship between the
occurrence of reverberation/streak artifact and the size (width or height)
of the urocystolith or the ultrasonographic frequency. There was, however,
a statistically relevant relationship between ultrasonographic imaging freq
uency and the occurrence of acoustic shadowing and between urocystolith hei
ght (thickness) and the occurrence of acoustic shadowing. However, regardle
ss of ultrasound frequency, acoustic shadowing was observed less than 35% o
f the time in any of the urocystolith mineral types examined. Based on the
imaging of the bladder phantom supporting apparatus (7.0 mm bolts covered b
y plastic), the accurate characterization of a curved object surface direct
ly facing the transducer was found to be directly related to the frequency
of sound used for imaging and at best predictably limited to curved vs flat
, Accurate measurement of the maximum transverse dimension of an echogenic
curved object or accurate characterization of the lateral borders of such a
n object was considered unlikely with general ultrasonographic equipment of
the frequencies studied. Therefore, detailed architectural characterizatio
n of urocystoliths suitable for mineral composition prediction is considere
d highly unlikely with general pulse-echo ultrasonographic techniques.