Introduction: It is important to know whether a parotid gland lesion is in the superficial or deep lobe for preoperative planning. We aimed to investigate the ability of 3.0 T magnetic resonance (MR) imaging with surface coils to identify the intraparotid facial nerve and locate parotid gland lesions, in comparison to other indirect landmark methods.
Methods: We retrospectively evaluated 50 consecutive patients with primary parotid gland lesions. The position of the facial nerve was determined by tracing the nerve in the stylomastoid foramen and then following it on sequential MR sections through the parotid gland. The retromandibular vein and the facial nerve line (FN line) were also identified. For each radiologist and each method, we determined the diagnostic ability for deep lobe lesions and superficial lobe lesions, as well as accuracy. These abilities were compared among the three methods using the Chi-square test with Yates' correction.
Results: Mean diagnostic ability for deep lobe lesions, the diagnostic ability for superficial lobe lesions, and accuracy were 92%, 86%, 87%, respectively, for the direct identification method; 67%, 89%, 86%, respectively, for the retromandibular vein method; and 25%, 99%, 90% , respectively, for the FN line method. The direct identification method had significantly higher diagnostic ability for deep lesions than the FN line method (P < 0.01), but significantly lower diagnostic ability for superficial lobe lesions than the FN line method (P < 0.01).
Conclusion: Direct identification of the intraparotid facial nerve enables parotid gland lesions to be correctly located, particularly those in the deep lobes.