Detection of intraocular foreign body with ultrasound

Objective: To evaluate the accuracy of emergency ultrasound in the diagnosis of varying location, size, and composition (metal, plastic, wooden) of intraocular foreign body in a sheep model. Methods: This was an experimental study using freshly harvested adult articulated sheep eyes. A total of 8 fresh sheep heads were used. Two of the eyes were randomised to include more than one foreign body, the first one with three different sizes of metallic fragments of 0.46x0.81x0.46 mm, 0.46x0.93x0.46 mm, 0.46x0.82x0.46 mm and the second eye containing plastic fragments with a size of 0.97x0.67x0.67 mm and 1.00x0.67x0.67 mm. The wooden fragment was 0.52x0.76 mm and the metallic fragment was 0.45x0.46x0.46 mm. Half of the eyes were chosen as the control group and the sclera of the control group was also punctured. A total of 5 residents, from different postgraduate years, and 2 attending physicians participated in the study. Results: Regardless of the nature and number of foreign bodies, correct identification of the intraocular foreign body was achieved in 44 of the 56 assessments. The overall sensitivity was 78% (95% confidence interval [CI], 65%-87%), and the overall specificity was 30% (95% CI, 19%-44%). Positive and negative predictive values were 53% and 58% respectively. Conclusion: The dynamic nature of the ultrasound examination may be helpful in localising and determining the size and composition of the foreign body, but if the intraocular foreign body is a concern and the ultrasound and physical examination are negative, another imaging modality, such as a computed tomography scan of the globe, should be obtained.