Three porcine cardiopulmonary blocks were perfused, preserved at 4°C for 6 h and reconditioned. Lungs were randomly seeded with different patterns of echogenicity target nodules (9 water balls, 10 fat, and 11 muscles; total n = 30). Three micro-convex US probes were assessed in an open setting on the pleural surface: PROBE 1, endobronchial US 5–10 MHz; PROBE 2, laparoscopic 4–13 MHz; PROBE 3, fingertip micro-convex probe 5–10 MHz. US probes were evaluated regarding the number of nodules localized/not localized, the correlation between US and open specimen measurements, and imaging quality.
For detecting target nodules, the sensitivity was 100% for PROBE 1, 86.6% for PROBE 2, and 78.1% for PROBE 3. A closer correlation between US and open specimen measurements of target diameter (r = 0.87; P = 0.0001) and intrapulmonary depth (r = 0.97; P = 0.0001) was calculated for PROBE 1 than for PROBES 2 and 3. The imaging quality was significantly higher for PROBE 1 than for PROBES 2 and 3 (P < 0.04).
US examination with micro-convex probes to detect pulmonary nodules is feasible in an ex vivo lung perfusion model. PROBE 1 achieved the best performance. Clinical research with the endobronchial US micro-convex probe during minimally invasive thoracic surgery is advisable.