Introduction

Facial nerve which is the VII of cranial nerves and part of the facial nerve transverses the parotid gland where the parotid gland tumors maybe ocurr. Thus,intraoperative injury may cause irreversible consequence that affect the quality of life seriously. However, the conventional MRI is unable to display the facial nerve in the parotid gland accurately^{1, 2}. It is important to find an effective method to display the facial nerve. T2WI-3D-FFE sequence is a three-dimensional high-resolution T2-weighted sequence, which can inhibit the fat background and blood flow signals, and has significant advantages in peripheral nerve. In this study, we use T2WI-3D-FFE sequence to display the facial nerve and explore the localization with parotid tumors.

Methods

Thirty patients with solitary tumor of parotid glands (27 benign tumors and 3 malignant tumors) were recruited from the First Affiliated Hospital of Dalian Medical University. All patients were scanned using a 3.0 T MR scanner (Ingenia CX, Philips Healthcare, the Netherlands) with a 32 channel phase-array head coil. The MR protocol is T2WI-3D-FFE sequence. TR/TE = 8.3/4.1 ms, flip angle=30, FOV=220×220×65 mm³, Voxel=0.65mm×0.65mm×1.00mm, matrix=340×339×130, slice thickness=1.0mm, Gap=-0.5. The serial images were reconstructed in multiplanar reconstruction (MPR) and curvilinear planar reconstruction (CPR). Altogether there were 30 facial nerves on the tumor side and the contralateral sides were measured as the control group. Two radiologists (with 3 and 5 years radiology experiences respectively) independently scored the display of the facial nerve in T2WI-3D-FFE images according to Previous work ^{3, 4}. ROI was placed on the nerve which is 1 mm far from the stylomastoid foramen and the place where the facial nerve enters the parotid gland. At the same time, the signal value of parotid parenchyma and image background were measured in the same slice. The signal intensity ratio (SIRN) is the comparison between the signal of facial nerve and the parotid gland in the same slice⁵. The nerve length of the mastoid and parotid gland is measured along the course of the reconstructed images. The tumor locations were categorized as deep or superficial lobes on the basis of observing the facial nerve on MRI and tumor directly. The other 3 indirect methods: the facial nerve line (FNL), retromandibular vein (RMV), and Utrecht line (UL) were also be recorded. The true positions of parotid tumors were confirmed by surgery. All the measurement data are expressed by mean ± standard deviation. The diagnostic accuracy, sensitivity and specificity for localizing parotid lesions using each method were calculated and compared using the McNemar tests. The scores of the image quality were evaluated on paired (mastoid and parotid gland) data by using a Wilcoxon test. For SNR and CNR, paired t testing was used to assess facial nerve (mastoid and parotid gland) differences. The interobserver reliability on qualitative evaluation was assessed via intraclass correlation coefficient (excellent agreement if ICC > 0.8; good agreement if ICC > 0.6).

Results

Measurement consistency between the two observers was good (ICC＞0.6, p<0.05). The average image quality scores are satisfied, SNR and CNR of the intraparotid facial nerve are better than the mastoid segment (Table 1). The SIRN and length of the mastoid and parotid gland segments are shown in Table 2. The direct method means judging the deep or superficial lobe by manifestation the facial nerve directly. It showed the highest sensitivity and specificity for localization of deep lobe parotid gland tumors (85.7%,100% respectively) (Table 3).

Discussion

In this study, T2-3D-FFE sequence can show the facial nerve of the parotid gland segments with high signal and quality. Moreover, compared with other indirect methods, the direct method showed the highest sensitivity and specificity for localization of deep lobe parotid gland tumors. Although the direct method has no statistical differences, which may be due to the small sample size, direct display of the course of the facial nerve can help surgeons make appropriate surgery plan pre-operation. T2-3D-FFE is a sequence of stimulation echoes that outline the surrounding nerve structure through the water in the nerve, and combined with the fat inhibition sequence, the nerves are high signal, clearly showing the neural section in the soft tissue, thus, the parotid segments of the facial nerve can be significantly contrasted with the surrounding parotid tissue⁶.

Conclusion

T2-3D-FFE is able to visualize the intraparotid facial nerves which can not be demonstrated on conventional MR imaging. It also provides detailed morphological information on the nerve relative to adjacent structures preoperatively.

Acknowledgements

No acknowledgement found.