**2. Materials and Methods**

A total of 20 adult volunteers (9 males and 11 females, mean age 41.5 years, age range 18–56 years) and 19 elderly volunteers (8 males and 11 females, mean age 67.8 years, age range 60–80 years) over the age of 60 years, with no sublingual gland-related diseases, as confirmed by both a history and clinical examination, were recruited (Table 1). In addition, 7 consecutive patients (3 males and 4 females, mean age 43.4 years, age range 19–76 years) were also recruited, with 5 having inflammations of the oral floor, including the sublingual glands, and 2 with ranulas (Table 1). The image of a single side (randomly chosen) or a disease-related side of the sublingual gland ducts was used, since only single images could be acquired at one given time for functional evaluation. The total volume of the sublingual gland ducts was also analyzed using the images. Approval for the present study was obtained from the institutional review board of Kyushu Dental University (No. 20-27).


**Table 1.** Subjects.

As in our previous reports, all images were acquired using a 1.5T full-body MR system (EXCELART Vantage powered by Atlas PPP; Toshiba, Tokyo, Japan) with a head coil (Atlas Head SPEEDER) to visualize the sublingual gland ducts, such as the parotid and submandibular gland ducts, according to Oda et al. [14]. T1-weighted, short tau inversion recovery (STIR), three-dimensional (3D) fast asymmetric spin-echo, and 2D-FASE images were acquired for each subject. The MRI parameters that were used are shown in Table 2. The 2D-FASE images were acquired after a single excitation with specific encoding for each echo. Fat saturation suppressed signals from subcutaneous fat.

The 3D MR sialography for sublingual gland ducts was performed as described by Oda et al. [14]. Briefly, in the same session where conventional MR studies of the sublingual glands were obtained, MR sialography was performed using 3D-FASE sequencing. In the 3D-FASE imaging, after a single excitation, images were acquired with a specific encoding for each echo. Fat saturation suppressed the signals from the subcutaneous fat. The imaging volume was centered parasagittally for the midline of the sublingual gland. In all volunteers and patients, post-processing of the MR sialographic images was performed for maximum intensity projection (MIP) reconstructions. Since 3D acquisitions can be reformatted into any required orientation, the sublingual gland ducts were identified on an initial set of axial 3D-FASE images, and oblique sagittal acquisition was used to capture the image of the parotid gland and/or submandibular gland ducts. The imaging time required for MR sialographic 3D reconstruction images using 3D-FASE sequencing was less than 5 min.


**Table 2.** Imaging parameters of each sequence.

TR: Repetition time, TE: Echo time, FOV: Field of view, STIR: Short T1 inversion recovery, T1WI: T1-weighted image, 2D-FASE: 2-dimensional fast asymmetric spin-echo, 3D-FASE: 3-dimensional fast asymmetric spin-echo.

Dynamic MR sialographic images and data were acquired using the method described by Oda et al. [14]. First, 2D-FASE sequencing was repeated every 18 s of acquisition time and 12 s of interval time before and after the placement of several drops of 5% citric acid (1 mL) on the tongue, using a device similar to a syringe to acquire the dynamic MR sialography. Fat saturation was also applied for the suppression of signals from subcutaneous fat. The acquisition time of the dynamic MR sialography was about 7 min after stimulation. For the prevention of movement artifacts, head rests were used with a flat long cord with non-magnetic materials.

Each digitized image acquired by the dynamic MR sialography was linked to the Ziostation2 (Ziosoft, Tokyo, Japan). The detectable area in the parotid or submandibular gland ducts on the respective images and the time from post-stimulation to the return to the baseline state of the ducts pre-stimulation were measured using the scanner-computer analysis system. For each patient, the change in ratio of the detectable area in the sublingual gland ducts in respective images to the detectable area pre-citric acid stimulation was also analyzed. A graph demonstrated the connection between the time post-stimulation (*x*-axis) and the change ratio of the dynamic MR sialographic data (*y*-axis). We commonly used the graph to show the connection between the time, post-stimulation, and the change ratio for the standardization of volunteers and patients.

Using the graph of dynamic MR sialography, the diagnostic parameters were also analyzed as follows: (1) the maximum area of the sublingual gland ducts pre-citric acid stimulation; (2) the change ratio (change ratio = detectable area of sublingual gland ducts post-citric acid stimulation/detectable area pre-citric acid stimulation); (3) the time from the end of post-stimulation to the occurrence of the maximum area of the sublingual gland ducts; and (4) the time it took for the sublingual gland ducts to decrease from their maximum level to 50% of the pre-stimulation level.

The Mann–Whitney U test was used to examine the differences between the following: (1) the maximum area of the sublingual gland ducts between the adult and elderly volunteers; (2) the degree of difference between the maximum and minimum duct areas based on computer calculations between the adult and elderly volunteers; (3) the time from the end of citric acid stimulation to the occurrence of the maximum area of the sublingual gland ducts between the adult and elderly volunteers; and (4) the time required for the sublingual gland ducts to decrease from their maximum level to the 50% pre-citric acid stimulation level between the adult and elderly volunteers. *p* values less than 0.05 indicated a significant difference.
