to distinguish benign from malignant salivary neoplasms (Keyes et al. 1994). The variable uptake of FDG by pleomorphic adenomas and the increased uptake and SUVs by Warthin tumors result in significant false positives (Jeong et al. 2007; Roh et al. 2007). In a similar manner, adenoid cystic carcinomas, which are relatively slower growing, may not accumulate significant concentrations of FDG and demonstrate low SUVs and therefore contribute to the false negatives (Jeong et al. 2007; Keyes et al. 1994). False negatives may also be caused by the relatively lower mean SUV of salivary tumors (SUV 3.8 ± 2.1) relative to squamous cell carcinoma (SUV 7.5 ± 3.4) (Roh et al. 2007). The low SUV of salivary neoplasms may also be obscured by the normal uptake of FDG by salivary glands (Roh et al. 2007). In general, FDG PET has demonstrated that lower grade malignancies tend to have lower SUV and vice versa for higher grade malignancies (Jeong et al. 2007; Roh et al. 2007). FDG PET has been shown to be more sensitive and specific compared to conventional CT or MRI (Otsuka et al. 2005; Cermik et al. 2007; Roh et al. 2007). Small tumor size can contribute to false negative results and inflammatory changes contribute to false positive results (Roh et al. 2007). The use of concurrent salivary scintigraphy with 99mTc‐pertechnetate imaging can improve the false positive rate by identifying Warthin's tumors and oncocytomas, which tend to accumulate pertechnetate (and retain it after induced salivary gland washout) and have increased uptake of FDG (Uchida et al. 2005).
Diagnostic Imaging Anatomy
PAROTID GLAND
The average adult parotid gland measures 3.4 cm in AP, 3.7 cm in LR and 5.8 cm in SI dimensions and is the largest salivary gland. The parotid gland is positioned high in the suprahyoid neck directly inferior to the external auditory canal (EAC) and wedged between the posterior border of the mandible and anterior border of the styloid process, sternocleidomastoid muscle, and posterior belly of the digastric muscle (Figures 2.17 through 2.19 ; 2.30 through 2.36). This position, as well as the seventh cranial nerve which traverses the gland, divides the gland functionally (not anatomically) into superficial and deep “lobes.” Its inferior extent is to the level of the angle of the mandible where its “tail” is interposed between the platysma superficially and the sternocleidomastoid muscle (SCM) deep to the tail of the parotid. The parotid gland is surrounded by the superficial layer of the deep cervical fascia. The parotid space is bordered medially by the parapharyngeal space (PPS), the carotid space (CS), and the posterior belly of the digastric muscle. The anterior border is made up of the angle and ramus of the mandible along with the masticator space (MS). The posterior border is made up of the styloid and mastoid processes and the SCM. The gland traverses the stylomandibular tunnel that is formed by the posterior border of the mandibular ramus, the anterior border of the sternocleidomastoid muscle, the anterior border of the stylomandibular ligament, and the anterior border of the posterior belly of the digastric muscle and the skull base on its superior aspect (Som and Curtin 1996; Beale and Madani 2006). The external carotid artery (ECA) and retromandibular vein (RMV) traverse the gland in a craniocaudal direction, posterior to the posterior border of the mandibular ramus. The seventh cranial nerve (CN VII) traverses the gland in the slightly oblique anteroposterior direction from the stylomastoid foramen to the anterior border of the gland passing just lateral to the RMV. The seventh cranial nerve divides into five branches (temporal, zygomatic, buccal, mandibular, and cervical) within the substance of the gland. Prior to entering the substance of the parotid gland, the facial nerve gives off small branches, the posterior auricular, posterior digastric, and the stylohyoid nerves. The intraparotid facial nerve and duct can be demonstrated by MRI using surface coils and high‐resolution acquisition (Takahashi et al. 2005). Because the parotid gland encapsulates later in development than other salivary gland, lymph nodes become incorporated into the substance of the gland. The parotid duct emanates from the superficial anterior part of the gland and is positioned along the superficial surface of the masseter muscle. Along the anterior aspect of the masseter muscle, the duct turns medially, posterior to the zygomaticus major and minor muscle to penetrate the buccinator muscle and terminates in the oral mucosa lateral to the maxillary second molar. Around 15–20% of the general population also has an accessory parotid gland which lies along the surface of the masseter muscle in the path of the parotid duct.
Figure 2.30. Axial CT of the neck demonstrates the intermediate to low density of the parotid gland.
Figure 2.31. Reformatted coronal CT of the neck at the level of the parotid gland demonstrating its relationship to adjacent structures. Note the distinct soft‐tissue anatomy below the skull base.
Figure 2.32. Reformatted sagittal CT of the neck at the level of the parotid gland demonstrating its relationship to adjacent structures including the external auditory canal. Note the slightly denser soft‐tissue density in the parotid tail, the so‐called “earring lesion” of the parotid gland. Cervical lymphadenopathy (arrow) was diagnosed at surgery.
Figure 2.33. Axial T1 MRI image at the level of the parotid gland demonstrating the slightly higher signal as compared to skeletal muscle but less than subcutaneous fat.
Figure 2.34. Coronal STIR MRI image at the level of the parotid gland demonstrating the nulling of the subcutaneous fat signal on STIR images and low signal from the partially fatty parotid gland.
Figure 2.35. Sagittal fat suppressed T1 MRI image of the parotid gland demonstrating mild enhancement and lack of subcutaneous fat signal in the upper neck but incomplete fat suppression at the base of the neck.
Figure 2.36. Axial CT scan (a) and corresponding PET scan (b) at the level of the parotid gland. Note the asymmetric slightly higher uptake on the right corresponding to partially resected parotid gland on the left, confirmed by CT.
In the pediatric population, the parotid gland is isodense to skeletal muscle by CT and becomes progressively but variably fatty replaced with aging. The CT density will therefore progressively decrease over time (Drumond 1995). By MRI, the parotid gland is isointense