Comparing the double-echo steady-state with water excitation and constructive interference in steady state sequence techniques for identifying extracranial facial nerve and tumor positions in patients with parotid tumors

ABSTRACT

BACKGROUND AND PURPOSE:

Reliable preoperative visualization of facial nerve morphology and understanding the spatial relationship between the facial nerve and tumor in the parotid gland can help clinicians perform safe and effective surgeries. Hence, this study aimed to compare the image quality of extracranial facial nerves obtained using double-echo steady-state with water excitation (DESS-WE) and constructive interference in steady state (CISS) sequences and evaluate their diagnostic efficacy in the localization of parotid tumors.

MATERIALS AND METHODS:

In total, 32 facial nerves of 16 healthy volunteers and 25 facial nerves of 25 patients with parotid tumors were included in this retrospective study. All participants underwent noncontrast-enhanced extracranial facial nerve magnetic resonance imaging with DESS-WE and CISS with a 3T MR scanner equipped with a 64-channel head and neck coil. Image quality was subjectively evaluated using a 5-point Likert scale by two radiologists. Inter- and intra-rater agreements were assessed using the Cohen kappa coefficient (κ). Receiver operating characteristic analysis was performed, and the diagnostic efficacies of DESS-WE and CISS images in localizing parotid tumors were calculated.

RESULTS:

For healthy volunteers (11 men and 5 women; median age, 26 years), image quality scores for CISS were significantly higher than those for DESS-WE for the discrimination of the temporofacial and cervicofacial trunks (both, p <0.001). In patients with parotid tumors (12 men and 13 women; median age, 58 years), CISS performed better than DESS-WE in terms of visualizing the spatial relationship of the facial nerve to the tumor and diagnostic confidence (both, p<0.001). Regarding the localization of parotid tumors, CISS showed excellent performance, comparable to that of DESS-WE (area under the curve, 0.981 versus 0.942, p = 0.1489).

CONCLUSIONS:

CISS achieved diagnostic performance comparable to DESS-WE in parotid tumor localization, with favorable image quality and more reliable morphological visualization of the facial nerve. ABBREVIATIONS: 3D = three-dimensional; CISS = constructive interference in steady state; DESS-WE = double-echo steady-state with water excitation; IQS = image quality score; AUC = area under the curve.