Radiological assessment of the fossa of Rosenmüller using cone-beam computed tomography: A retrospective study

Abstract

BACKGROUND

The fossa of Rosenmüller (RF) is a herniation of nasopharyngeal mucosa through a defect between the skull base and the superior-most fibers of the superior constrictor muscle. Its anatomy and relationships help explain symptoms commonly seen in nasopharyngeal disorders, as well as potential pathways for direct disease extension. This retrospective study aims to assess RF anatomy using two- and three-dimensional measurements and to evaluate these findings according to demographic parameters.

AIM

To assess the anatomy of RF via two- and three-dimensional measurements and evaluate the findings according to demographic parameters.

METHODS

Cone-beam computed tomography images from 500 Turkish patients were retrospectively assessed by two observers, each performing measurements twice. Images were acquired using a Planmeca Promax 3D Max system (Planmeca, Helsinki, Finland) (70-90 kVp; 9-10 mA; exposure time 10-15 seconds). Measurements included RF depth, width, angle, and volume. Patient age and sex were recorded. Volume measurements were performed using 3D Doctor software (Able Software Corp., Lexington, United States). The Mann-Whitney U test was used to assess differences between groups. Intraclass correlation coefficients were calculated to evaluate intra- and inter-observer agreement. Statistical significance was set at P < 0.05.

RESULTS

The mean width was 4.21 ± 1.43 mm (median: 4.16; range: 1.17-11.09), and the mean depth was 12.82 ± 3.39 mm (median: 12.84; range: 4.79-25.86). The mean angle relative to the sagittal plane was 52.34° ± 9.11° (median: 51.40; range: 9.28-79.90). Volumetric analysis demonstrated a mean RF volume of 334.78 ± 329.93 mm³ (median: 210.67; range: 29.31-1632.76), indicating substantial interindividual variability. No statistically significant sex-related differences were identified in width (female: 4.20 ± 1.46 mm; male: 4.24 ± 1.41 mm; P = 0.885), depth (female: 12.82 ± 3.39 mm; male: 12.80 ± 3.44 mm; P = 0.973), or angle (female: 51.66° ± 9.75°; male: 53.52° ± 7.88°; P = 0.296). Similarly, volumetric measurements did not differ significantly between females and males (307.26 ± 304.78 mm³ vs 383.26 ± 368.99 mm³, respectively; P = 0.336). Intraclass correlation coefficients demonstrated excellent intra- and inter-observer agreement for all measurements (P < 0.001).

CONCLUSION

Cone-beam computed tomography may provide useful information for assessing the RF.

Key Words: Dentomaxillofacial radiology; Rosenmuller fossa; Cone-beam computed tomography; Radiological assessment; Measurement

Core Tip: The fossa of Rosenmüller is a key anatomical region of the nasopharynx and represents the most frequent site of origin of nasopharyngeal carcinoma. Despite its clinical significance, detailed three-dimensional morphometric data on this structure remain limited. In the present study, cone-beam computed tomography enabled accurate three-dimensional linear and volumetric evaluation of the fossa of Rosenmüller. The combined use of volumetric, surface area, and linear measurements provides a more comprehensive characterization of its complex morphology. These findings offer normative reference data and support the use of cone-beam computed tomography as a valuable complementary imaging modality for nasopharyngeal anatomical and pathological assessment.

INTRODUCTION

The fossa of Rosenmüller (RF), also known as the pharyngeal recess, is a bilateral indentation of the lateral nasopharyngeal wall located posterior to the cartilaginous portion of the Eustachian tube and the torus tubarius[1]. It represents a herniation of nasopharyngeal mucosa through a defect between the skull base and the superior-most fibers of the superior constrictor muscle. Its anatomy and relationships help explain symptoms commonly seen in nasopharyngeal disorders, as well as pathways for direct disease extension[2].

The RF is of particular clinical importance because it is the most common site of origin for nasopharyngeal carcinoma (NPC), a malignancy with often subtle early radiographic features that are critical to detect for optimal therapeutic outcomes[3]. This retrospective study aims to assess RF anatomy using two- and three-dimensional measurements and to evaluate these findings according to demographic parameters.

MATERIALS AND METHODS

Approval for the present study was obtained from the Ethics Committee of the Faculty of Dentistry, Ankara University, Ankara, Türkiye, and the study was conducted in accordance with the Declaration of Helsinki. Cone-beam computed tomography (CBCT) images of 500 Turkish patients aged 18-60 years were retrospectively selected from the digital database of Ankara University.

All scans were acquired using a ProMax 3D Max CBCT scanner (Planmeca, Helsinki, Finland) equipped with a flat-panel detector. Imaging parameters were set at 96 kVp and 1-8 mA, with a field of view of 230 mm × 160 mm and a voxel size of 0.4 mm. The exposure time ranged between 5.087 and 5.166 seconds, and the dose-area product was 46 mGy·cm². Images with motion artifacts, prior surgical intervention, or pathological findings affecting the nasopharyngeal region were excluded.

RF measurements included depth, orifice width, angle relative to the sagittal plane, and volume. Patient age and sex were also recorded. Linear measurements were performed on axial CBCT sections. All images were evaluated in a dimly lit room using a 15.6-inch laptop monitor (Qosmio F75-5-3D350, Toshiba, Tokyo, Japan) with a resolution of 1920 pixels × 1080 pixels and 32-bit color depth.

For volumetric analysis, axial CBCT datasets were exported in DICOM format and imported into third-party software, 3D Doctor (Able Software Corp., Lexington, United States), which utilizes vector-based segmentation technology. On axial images, fossa width, depth, angle, and volume were measured and recorded (Figure 1A and B).

Figure 1 Measurement of the width, depth, angle and volume of the fossa of Rosenmüller. A: Width and depth measurement. The red short line depicts the width of the orifice of the fossa of Rosenmüller, whereas the red long straight line measured from the midpoint of the width shows the depth; B: Angle measurement; C: Volumetric measurement.

For volumetric measurements, CBCT images were manually segmented on consecutive axial slices, and three-dimensional reconstruction was automatically generated by the software, enabling apicocoronal visualization and volumetric calculation (Figure 1C). All linear and volumetric measurements were performed using the same software by two experienced dento-maxillofacial radiology specialists at two time points, separated by a 2-week interval. A subset of 50 images was reevaluated to assess intra-observer agreement.

The Mann-Whitney U test was used to assess differences between groups. Intraclass correlation coefficient (ICC) values were calculated to evaluate intra- and inter-observer agreement. Statistical significance was set at P < 0.05.

RESULTS

Overall descriptive statistics for RF morphology are presented in Table 1. The mean width was 4.21 ± 1.43 mm (median: 4.16; range: 1.17-11.09), and the mean depth was 12.82 ± 3.39 mm (median: 12.84; range: 4.79-25.86). The mean angle relative to the sagittal plane was 52.34° ± 9.11° (median: 51.40; range: 9.28-79.90). Volumetric analysis demonstrated a mean RF volume of 334.78 ± 329.93 mm³ (median: 210.67; range: 29.31-1632.76), indicating substantial interindividual variability (Table 1).

Table 1 Descriptive statistics of the morphometric measurements of the fossa of Rosenmüller (n = 1000 sides).

Variable	n	mean	SD	Median	Minimum	Maximum
Width (mm)	1000	4.2134	1.43496	4.1615	1.17	11.09
Depth (mm)	1000	12.8159	3.39217	12.8350	4.79	25.86
Angle (°)	1000	52.3415	9.11407	51.3950	9.28	79.90
Volume (mm³)	1000	334.7796	329.92826	210.6730	29.31	1632.76

A total of 500 Turkish patients (18-60 years) were included in the study. The mean age was 31.62 ± 9.50 years, as summarized in Table 2. Bilateral RF measurements were performed, yielding 1000 sides for morphometric evaluation.

Table 2 Demographic characteristics of the study population.

Variable	N	Minimum	Maximum	mean	SD
Age (years)	500	18.00	60.00	31.6207	9.50314

Mean and SD with minimum and maximum values.

Gender-based comparisons are provided in Table 3. The female group comprised 315 patients (630 sides), and the male group comprised 185 patients (370 sides). No statistically significant sex-related differences were identified in width (female: 4.20 ± 1.46 mm; male: 4.24 ± 1.41 mm; P = 0.885), depth (female: 12.82 ± 3.39 mm; male: 12.80 ± 3.44 mm; P = 0.973), or angle (female: 51.66° ± 9.75°; male: 53.52° ± 7.88°; P = 0.296). Similarly, volumetric measurements did not differ significantly between females and males (307.26 ± 304.78 mm³ vs 383.26 ± 368.99 mm³, respectively; P = 0.336), as shown in Table 3.

Table 3 Comparison of fossa of Rosenmüller measurements between sexes, mean ± SD.

Parameter	Female (n = 630)	Male (n = 370)	P value
Width (mm)	4.1988 ± 1.45850	4.2392 ± 1.40960	0.885
Depth (mm)	12.8239 ± 3.39043	12.8018 ± 3.43630	0.973
Angle (°)	51.6567 ± 9.75115	53.5153 ± 7.87686	0.296
Volume (mm³)	307.2620 ± 304.78089	383.2630 ± 368.98751	0.336

P values indicate between-group comparisons; statistical significance was set at P < 0.05.

Collectively, these findings establish normative CBCT-based linear and volumetric reference values for the RF within the studied population and suggest that the evaluated morphometric parameters are not significantly influenced by sex (Tables 1-3).

Table 4 presents ICC values for intra- and inter-observer agreement for width, depth, angle, and volume measurements. Both intra- and inter-observer agreement were excellent.

Table 4 Intraclass correlation coefficient values for intra-observer and inter-observer agreement for all measurements.

Parameter	ICC	P value	95% confidence interval
Intraobserver observer 1/observer 2
Width	0.971/0.965	< 0.001	0.941-0.939
			0.974-0.972
Depth	0.973/0.954	< 0.001	0.967-0.965
			0.995-0.996
Angle	0.965/0.961	< 0.001	0.951-0.964; 0.997-0.995
Volume	0.952/0.955		0.941-0.939; 0.991-0.989
Interobserver observer 1-2
Width	0.918	< 0.001	0.901-0.943
Depth	0.912	< 0.001	0.898-0.931
Angle	0.909	< 0.001	0.901-0.913
Volume	0.908	< 0.001	0.889-0.911

ICC: Intraclass correlation coefficient.

DISCUSSION

The RF is a complex anatomical recess of the nasopharynx and represents the most frequent site of origin of NPC. Despite its well-established clinical relevance, quantitative three-dimensional morphometric data on this region remain limited. In the present study, CBCT images of the RF were evaluated using linear, angular, and volumetric measurements, enabling comprehensive anatomical characterization. The excellent ICC values obtained indicate high intra- and inter-observer repeatability. Within the limitations of this study, these findings support the use of CBCT for RF assessment.

A previous CBCT study of posterior nasopharyngeal wall structures in a Turkish cohort reported lower mean RF depth values and identified statistically significant sex differences in RF width, along with an age-related increase in RF-posterior nasal spine distance[4]. In contrast, our analysis did not demonstrate significant sex differences in width, depth, angle, or volume. These discrepancies may reflect methodological heterogeneity, including differences in CBCT acquisition parameters (e.g., device and voxel size and, importantly, measurement definitions (purely linear axial metrics vs segmentation-based 3D assessment). Future studies that standardize RF boundary definitions and integrate skeletal reference distances with volumetric segmentation would improve cross-study comparability.

Previous investigations of the RF have primarily relied on linear morphometric measurements obtained from computed tomography (CT) or CBCT imaging. Kaplan et al[5] analyzed posterior nasopharyngeal wall structures using CBCT and demonstrated significant anatomical variability in recess depth and width. Similarly, Serindere et al[6] reported morphometric distances and anatomical relationships of the RF in a large CBCT cohort, emphasizing the clinical importance of this region. However, neither study incorporated volumetric or surface area analysis.

The present study expands upon these findings by integrating three-dimensional volumetric evaluation, providing a more realistic representation of this irregular anatomical cavity.

Linear morphometric measurements alone may not adequately reflect the true anatomical characteristics of irregular cavities such as the RF. Grauer et al[7] emphasized that volumetric airway analysis provides superior anatomical accuracy compared with two-dimensional measurements, particularly for complex three-dimensional structures. Our findings support this concept, as volumetric measurements revealed substantial interindividual variability that could not be fully captured by linear parameters alone.

Volumetric measurements in our population showed a wide range, reflecting notable anatomical heterogeneity. This variability underscores the need to establish population-specific normative reference values; without such data, distinguishing physiological variation from early pathological change becomes challenging. This is particularly important for the RF, as NPC often arises submucosally within this recess and may not produce obvious early radiologic contour changes[8,9].

No statistically significant sex-related differences were observed in our study. This contrasts with previous CBCT-based airway and nasopharyngeal studies reporting larger airway and recess dimensions in males[9,10]. Sexual dimorphism in craniofacial skeletal morphology, hormonal influences on growth patterns, and differences in muscular and connective tissue development may contribute to these variations. These findings suggest that RF assessment should consider sex-specific reference values to avoid misinterpretation.

Early detection of NPC remains a diagnostic challenge. King et al[8] demonstrated that early NPC often originates in the pharyngeal recess and may be overlooked without careful radiologic assessment. Quantitative morphometric evaluation may therefore enhance early diagnostic sensitivity by enabling detection of subtle asymmetry, focal volumetric expansion, or architectural distortion. The volumetric reference values established in this study may assist radiologists in identifying suspicious deviations from normal anatomy.

CBCT offers several advantages for morphometric evaluation of the nasopharyngeal region. Compared with conventional CT, CBCT provides isotropic voxel geometry, high spatial resolution, and reduced radiation dose, facilitating accurate three-dimensional reconstruction of complex anatomical structures[10,11].

Contrast-enhanced magnetic resonance imaging (MRI), following endoscopic evaluation, is the preferred modality due to its higher sensitivity for detecting invasion, lymphadenopathy, submucosal bulging, and adenoid enlargement. It has been suggested that asymptomatic individuals with positive MRI findings but negative endoscopic examination should undergo nasopharyngeal biopsy or close MRI surveillance. Imaging also helps guide the site and depth of biopsy. In contrast, CT provides superior evaluation of bony destruction[12].

The present study has several limitations that should be considered when interpreting the results. It is a single-center study with a sample limited to the Turkish population. The relatively limited soft-tissue contrast of CBCT may hinder visualization of mucosal detail or early infiltration. Additional limitations include the subjectivity of semi-automated manual segmentation and the absence of validation against a gold standard (e.g., cadaveric dissection or intraoperative measurements).

Although CBCT has lower soft-tissue contrast than MRI, its geometric accuracy makes it particularly well suited for anatomical morphometry. Therefore, CBCT should be considered a complementary modality to MRI in the assessment of nasopharyngeal anatomy.

CONCLUSION

The findings of this study also have implications for surgical and endoscopic procedures involving the nasopharynx. Accurate knowledge of RF morphology is essential for safe biopsy sampling, endoscopic navigation, and avoidance of adjacent vital structures, such as the internal carotid artery. Three-dimensional morphometric data may therefore contribute to improved procedural planning and reduced complication risk. CBCT imaging may provide valuable anatomical information for the evaluation of RF.