*3.2. Function of Sublingual Gland Ducts Evaluated by Dynamic MR Sialography*

Normal dynamic MR sialography images obtained before and after citric acid stimulation are shown in Figure 2. The sublingual gland ducts were identified as many bright, homogeneous, ascending linear structures in continuity with the sublingual glands, as mentioned above (Figure 2A). The many ducts became slightly clearer in a time-dependent fashion gradually after citric acid stimulation and up to 30–60 s post-stimulation. Thereafter, the many ducts became slightly clearer in a time-dependent fashion. In the graph demonstrating the relationship between the time course post-citric acid stimulation and the change ratio of the detectable area in the sublingual gland ducts, the area was seen at first to only increase slightly to 30 s in a time-dependent fashion (Figure 2B).

The volunteers' data are summarized in Table 4. Before citric acid stimulation, the maximum area of the sublingual gland ducts was 10.0 mm<sup>2</sup> (mean <sup>±</sup> SD = 10.0 <sup>±</sup> 4.6 mm2) in the 16 adult volunteers, 9.0 mm<sup>2</sup> (mean <sup>±</sup> SD = 9.0 <sup>±</sup> 3.4 mm2) in the 3 elderly volunteers, and 10.2 mm2 (mean <sup>±</sup> SD <sup>=</sup> 10.2 <sup>±</sup> 5.5 mm2) in the 5 patients (adult vs. elderly: *<sup>p</sup>* <sup>=</sup> 0.21, adult vs. patients: *<sup>p</sup>* <sup>=</sup> 0.46, elderly vs. patients: *p* = 0.92; Mann–Whitney U test). After citric acid stimulation, the maximum area of the parotid gland duct was 13.2 mm<sup>2</sup> (mean <sup>±</sup> SD <sup>=</sup> 13.2 <sup>±</sup> 5.3 mm2) in the 16 adult volunteers, 10.7 mm2 (mean <sup>±</sup> SD <sup>=</sup> 10.7 <sup>±</sup> 4.4 mm2) in the 12 elderly volunteers, and 11.0 mm2 (mean <sup>±</sup> SD <sup>=</sup> 11.0 <sup>±</sup> 5.4 mm2) in the 5 patients (adult vs. elderly: *p* = 0.53, adult vs. patients: *p* = 0.94, elderly vs. patients: *p* = 0.67; Mann–Whitney U test).

**Figure 2.** Dynamic MR sialographic images (**A**) and a graph (**B**) of sublingual gland ducts in a 26-year-old, healthy female volunteer. (**A**) The sublingual gland (arrowheads) and ducts (arrows) are gradually and slightly more clearly visualized after stimulation with citric acid for up to 60 s in a time-dependent manner. After 150 s, the many ducts became slightly clearer in a time-dependent fashion. (**B**) A graph of MR data of the sublingual gland ducts in Figure 2A demonstrates the connection between the time post-stimulation (*x*-axis) and the change ratio (*y*-axis). The area is seen at first to only increase slightly until 60 s in a time-dependent fashion. The maximum change ratio is about 1.2.

After stimulation, the time of occurrence of the maximum duct area varied from 30 s to 180 s in all subjects (mean ± SD = 62 ± 28 s in the 16 adult volunteers, mean ± SD = 63 ± 26 s in the 12 elderly volunteers, and mean ± SD = 54 ± 13 s in the 5 patients); (adult vs. elderly: *p* = 0.92, adult vs. patients: *p* = 0.40, elderly vs. patients: *p* = 0.39; Mann–Whitney U test). The time it took for the detectable duct area to return to almost 50% of its former area was about 115 s in all subjects (mean ± SD = 110 ± 39 s in the 12 adult volunteers, mean ± SD = 117 ± 57 s in the 12 elderly volunteers, and mean ± SD = 114 ± 25 s in the 5 patients); (adult vs. elderly: *p* = 0.76, adult vs. patients: *p* = 0.82, elderly vs. patients: *p* = 0.89; Mann–Whitney U test). No significant differences in the four parameters, including the change ratio

(adult vs. elderly: *p* = 0.24, adult vs. patients: *p* = 0.19, elderly vs. patients: *p* = 0.26; Mann–Whitney U test) in the detectable duct area of the sublingual glands were found among the three groups (Table 4).


**Table 4.** Summary of physical and dynamic MR sialographic data.

*3.3. Clinical Application of MR Sialography for Patients with Sublingual Gland Diseases*

In a 76-year-old man with inflammation of the right oral floor, many sublingual gland ducts continued with the sublingual glands in STIR, T1-weighted images, and MR sialography (Figure 3A–C).

**Figure 3.** STIR (**A**), T1-weighted images (**B**), and MR sialography (**C**) of a 76-year-old man with inflammation of the right oral floor. The disappearance (arrows) of many sublingual gland ducts in continuity with the sublingual glands is visualized using MR sialography.

In a 30-year-old woman with a ranula on the left, the mass lesion was detected in continuity with the sublingual glands in STIR and T1-weighted images and was thus diagnosed as a ranula (Figure 4A,B). In addition, the mass was derived from one of the many sublingual gland ducts in images obtained using MR sialography (Figure 4C).

**Figure 4.** STIR (**A**), T1-weighted image (**B**), and MR sialography (**C**) of a 30-year-old woman with a ranula on the left. The mass lesion is seen in continuity with the sublingual glands in STIR (**A**) and T1-weighted images (**B**) and is diagnosed as a ranula. The mass is derived from one of many sublingual gland ducts (arrow) (**C**).

#### **4. Discussion**

The most interesting result of the present study is that it is the first to show the imaging characteristics of the sublingual gland ducts obtained by 3D MR sialography. It is otherwise difficult to visualize the very thin and very short ducts, with a diameter and length of only 1 mm, as shown in anatomy textbooks [13]. This is a very significant first success in salivary gland imaging. The extraglandular portions of the typical sublingual gland ducts appeared as many bright, homogeneous, ascending linear structures in continuity with the sublingual glands. The figures of the sublingual gland ducts obtained by MR sialography were the same as those in a textbook of oral anatomy [13]. Therefore, the images of the structures could be confirmed to be sublingual gland ducts using MR sialography. However, the sublingual gland duct could not be detected in all subjects as the detection rate was about 57.1%. One possible explanation is that the sublingual gland ducts are so thin and short that they cannot be visualized in all subjects using MR sialography.

So far, the reason the sublingual gland ducts, with their thin and short size, have not been visualized before, even in MR images, is that visualization has been considered technically impossible. In addition, it was thought that there was little clinical significance in the visualization of sublingual gland ducts. However, these very sublingual gland duct-like structures were visualized in the MR sialography of patients with submandibular and/or parotid gland-related diseases. Therefore, we planned the present study of the visualization of sublingual gland ducts using MR sialography.

One other interesting result of the present study is that the visualization of sublingual gland ducts indicates the clinical significance of sublingual gland-related diseases. The visualization of sublingual gland ducts concretely demonstrated a mass derived from one of many sublingual gland ducts. Based on the imaging, the mass should have been diagnosed as a ranula. At the same time, the disappearance of many sublingual gland ducts in continuity with sublingual glands was visualized using MR sialography in patients with inflammation on the right. We would like to elucidate the clinical significance of the MR sialography of sublingual gland ducts for many kinds of diseases in the oral floor, including sublingual gland-related diseases.

One reason for this first success in the visualization of sublingual gland ducts is that MR systems, 3D computer vision, and image processing techniques have been fast advancing due to the growing computational power of current computer systems. Rapid advances in 3D data acquisition and post-processing technologies are expanding the potential applications of 3D displays. In the 1.5T full-body MR system (EXCELART Vantage powered by Atlas PPP; Toshiba, Tokyo, Japan) with a

head coil (Atlas Head SPEEDER), 3D-FASE was used for the sequencing of MR data sets through the acquisition of MR sialographic 3D reconstruction images, since this method was most likely to provide good resolution in a short period of time, as in our previous report [15]. That both types of images had good resolution using 3D-FASE sequencing may have been due to the section thickness being as little as 1 mm, despite the short acquisition time [16]. The section thickness was based on the advantage that 3D-FASE sequencing could excite a three-dimensional sample [17]. Thin slices produce images with good resolution while minimizing interference from partial volume effects and the formation of artifacts in the MR data and workstation. In addition, an adequate Fourier transform may be applied to 3D-FASE sequencing to produce high-resolution images [17]. Another advantage is that an image can be acquired using additional excitation, without conducting an additional imaging session, in cases when using a single excitation would not produce a satisfactory image. These unusual and useful characteristics of 3D-FASE sequencing allow for the avoidance of unnecessary exposure of patients to the RF pulse and an unnecessarily long acquisition time.

To our regret, little alteration of the dynamic curve was seen by dynamic MR sialography of sublingual gland ducts. This can be considered physiologically correct and reasonable [18,19]. Physiologically stimulated saliva production is the main role of the parotid glands [18,19]. The saliva flow rate of the parotid glands increases more quickly than that of the submandibular glands during citric acid stimulation [20–22]. Resting saliva is produced as the main role of the submandibular glands [18,19]. The salivary flow rate in the parotid gland during stimulation is twice as high as that in the rest phase, but less of an increase is found in the submandibular gland [23,24]. The main role of sublingual glands, however, is to keep the oral mucosa moist, but not to maintain resting and stimulated saliva flow. Therefore, little alteration of the dynamic curve via dynamic MR sialography of sublingual gland ducts was seen. We are now planning to elucidate the clinical significance of dynamic MR sialography of the sublingual gland ducts through its clinical application to many kinds of diseases in the oral floor, including sublingual gland-related diseases.

One possible limitation of the present data is the small sample size. The variables of race and sex could not be studied in this study sample. In addition, only a few clinical applications were examined. Therefore, further investigation is required. In the present study, there was little witness of movement artifacts by the volunteers, but we predicted that patients would move in MR examinations, despite our system preventing movement artifacts. Moreover, we paid attention to the possibility of visualizing sublingual gland ducts using dynamic MR sialography and its clinical application in the present study. Therefore, we could not elucidate the classification of the drainage of the sublingual glands in Bartholin's ducts and/or the duct of Rivinus. At the next stage, we should try to classify their drainage patterns. At the same time, we should try to elucidate how the presence of Bartholin's ducts may be related to ranula formation as the next stage. We added the sentence mentioned above in the revised manuscript.
