The effect of body wall on video signal analysis measurements

Rectangular body wall specimens were extracted from 16 juvenile swine and 9 adult beagle hounds after euthanasia. The body wall specimen included the epidermis to parietal membrane, with falciform fat removed. Ten images of a reference phantom with known attenuation and 10 additional images of the phantom with the specimen placed between the transducer and phantom surface were collected with a 5-MHz ultrasound system and computer with frame grabber board. Mean pixel values were converted to relative echogenicities. Echogenicity versus depth yielded an estimate of attenuation. An unpaired t test was applied to compare reference attenuation values with and without body wall, and a Pearson correlation was applied to body wall parameters versus measured attenuation through body wall. Measured attenuation through body wall increased significantly in dogs (P = 0.0016) and swine (P < 0.0001) when compared with phantom material alone. Increased attenuation positively correlated to body wall thickness (r = 0.6442) and mean gray level within body wall (r = 0.5069) for swine but not in canine. The presence of body wall in images used for video signal analysis significantly increases the measured attenuation in a phantom. This increase does not correlate with a measurable body wall parameter in dogs.