Ultrasound accurately identifies soft tissue foreign bodies in a live anesthetized porcine model

Some subcutaneous foreign bodies (FBs) are not easily visualized during physical examination and may not be detected on radiographic evaluation. Ultrasound (US) is capable of visualizing FBs of varying compositions. Previous studies have examined the use of US to detect FBs in deceased animal or human tissue. This study used live anesthetized porcine tissue to more closely model clinical conditions.

The objectives were to examine the test characteristics of US in the evaluation of FBs in living tissue and to evaluate if secondary findings such as surrounding edema and hematoma improve diagnostic accuracy.

Institutional Animal Care and Use Committee (IACUC) approval was obtained. FBs 1 cm in length and 1 to 3 mm in width were created from toothpicks (wood), 21-gauge needles (metal), and a broken ampule (glass) and inserted subcutaneously into an anesthetized 20-kg Yorkshire swine. There were 72 sites implanted with equal proportions of each FB type and null sites. Half of the FBs were inserted at time 0 and half were inserted after 2 hours. Immediately after placement, four blinded physicians performed US evaluations of the first 36 sites. At 2 hours after placement, they evaluated each of the original 36 sites and the 36 new sites. They documented the presence or absence of FBs and surrounding edema.

After initial FB placement, 122 of the 144 interpretations (85%) were correct, with a sensitivity of 85% (95% confidence interval [CI] = 79% to 92%) and a specificity of 86% (95% CI = 76% to 98%). No sites demonstrated surrounding edema. At 2 hours after placement, 127 of 144 interpretations (88%) for these same sites were correct, with a sensitivity of 87% (95% CI = 82% to 93%) and a specificity of 89% (95% CI = 81% to 97%). Of the 108 observations (27 sites that contained FBs), eight of the 108 (7%) observations had surrounding edema (four glass, three wood, one metal). For the 36 new sites with FBs placed 2 hours later, 126 of the 144 interpretations (83%) were correct, with a sensitivity of 88% (95% CI = 82% to 94%) and a specificity of 83% (95% CI = 73% to 95%). No sites had surrounding edema present.

Ultrasound was sensitive, specific, and accurate in identifying FBs in live anesthetized porcine tissue. Surrounding edema or hematoma 2 hours after placement was so infrequently observed that it was not possible to determine its influence on the test characteristics.