CT more accurately detects foreign bodies within the equine foot than MRI or digital radiography

摘要

Identification and characterization of foreign bodies in the distal limb of horses poses a diagnostic challenge. The aims of this prospective experimental cadaver study were to describe the appearance of five foreign body materials within the equine hoof using CT, MRI, and digital radiography (DR) and to compare interrater agreement among three reviewers. Fifty foreign bodies consisting of five materials were implanted at a solar location or a coronary location in 25 equine cadaver feet. The images were reviewed by three equine veterinarians experienced in advanced imaging interpretation, who were blinded to the material of the foreign body. Foreign bodies were graded on visibility and appearance. Sensitivity and specificity were calculated for accurate identification of the different materials. Interrater agreement was assessed using Fleiss' kappa. Computed tomography had higher visibility score, sensitivity/specificity, and interrater agreement for detection of all materials; particularly slate, glass, and dry wood, compared to the other imaging modalities. Soaked wood and plastic had lower sensitivity (31-33%) on CT with a similar attenuation of the two materials. Foreign bodies were often visible on MRI, although with similar appearance and unclear details. On DR, only slate and glass were visible. The interrater agreement for identifying the correct material was almost perfect for slate, glass, and dry wood (kappa = 0.92-1.00) and poor for plastic and soaked wood (kappa < 0.20) on CT. Interrater agreement was poor for all materials on MRI and DR (kappa < 0.20), with the except for fair (kappa = 0.28) for slate on DR and moderate (kappa = 0.28) for soaked wood on MRI.