Three-dimensional structure of the basal lamella of the middle turbinate

The middle turbinate’s basal lamella (3BL) is a variable landmark which needs to be understood in endoscopic transnasal skull base surgery. It comprises an anterior frontal and a posterior horizontal part and appears in its simplest depiction to be “L”-shaped, when viewed laterally. In this study we analyzed its 3D morphology and variations focusing on a precise and systematic description of the anatomy. CBCTs of 25 adults, 19 cadavers and 6 skulls (total: 100 sides) were investigated with the 3DSlicer software, creating 3D models of the 3BL. We introduced a novel geometrical classification of the 3BL’s shape, based on segments. We analyzed their parameters and relationship to neighboring structures. When viewed laterally, there was no consistent “L”-shaped appearance of the 3BL, as it is frequently quoted. A classification of 9 segment types was used to describe the 3BL. The 3BLs had in average of 2.95 ± 0.70 segments (median: 3), the most frequent was the horizontal plate (23.05% of all segments), next a concave/convex plate (22.71%), then a sigma plate (22.37%). Further types were rare. We identified a horizontal plate in 68% of all lateral views whilst 32% of the 3BLs were vertical. A sigma–concave/convex–horizontal trisegmental 3BL was the most common phenotype (27%). Globally, the sigma–concave/convex pattern was present in 42%. The 3BL adhered the ethmoidal bulla in 87%. The segmenting method is eligible to describe the 3BL’s sophisticated morphology.

www.nature.com/scientificreports/ and posterior ethmoidal cells ("struggle of the ethmoid") is associated with alterations of the 3BL's shape 7,10 . Lastly, each 3BL is unique 11 . Consequently, the identification of the 3BL and its optimal entry zone during ethmoidectomy can be challenging. However, there are hardly any studies which go beyond these statements. Most reviews only examine the 3BL on triplanar CT reconstructions. There is a lack of information about the 3BL's real 3D morphology, anatomical variations and their possible classification, posing the question if there is a most common phenotype. A precise assessment of the 3BL and its entry zone into the posterior ethmoid would also lead to a safer dissection. For these reasons, we performed a descriptive CT morphological study of the 3BL's 3D shape and anatomical features.

Methods
Case acquisition. In this retrospective study, non-enhanced cone beam CT scans (CBCT; slice thickness: 0.25 mm) of 25 anonymized patients (50 sides; 16 females and 9 males, mean age: 51.8 years, range: 25.6-83.6 years) performed for dental diagnostics were analyzed with the 3DSlicer software (https:// www. slicer. org) 12 . Pediatric cases as well as individuals with sinonasal pathologies (inflammation, polyps, tumor), trauma, surgery in the anamnesis were excluded. Furthermore, CBCTs of 19 formaldehyde-fixed cadaveric heads (38 sides) and 6 skulls (12 sides) provided by the Dept. of Anatomy, Semmelweis University, Budapest, Hungary were investigated. Here, same exclusion criteria were applied. A total of 100 sides have been analyzed.
Definition of the 3BL. The middle turbinate is composed of the lamina recurvata (the free turbinate part within the nasal cavity, usually understood as the middle turbinate) and the intraethmoidal lamina basalis (3BL). We postulated that the 3BL is a circumscribed object with defined borders (Fig. 1). The division relies mainly on the triplanar identification of neighboring structures: the 3BL is surrounded by ethmoidal cells and spaces, contrary to the MT, which is surrounded medially by the nasal cavity.
The medial border is the imaginary curve where the 3BL enters the nasal cavity and transitions into the lamina recurvata. The superior border of the 3BL is the attachment line on the ethmoidal roof. The lateral border is the curvy attachment line on the orbital lamina of the ethmoid bone (lamina papyracea), the medial wall of the maxillary sinus and eventually the medial surface of the palatine bone. The superior and lateral borders build together the attachment line on the skull base, orbit, maxilla and lateral nasal wall. We call this the superolateral attachment line (SLAL). Posteroinferiorly, the medial and lateral borders usually meet in an acute angle; it is mostly found at the region of the ethmoidal crest.
In each case, we created a 3D model of the 3BL using the Segment Editor tool of the 3DSlicer software. Furthermore, we analyzed the triplanar reconstructions. Each case has been revised on a second occasion.
Parameters of the 3BL. Triplanar CT reconstructions depict the 3BL in 2D from artificially chosen directions. We established a segmenting method to accurately and apparently describe the geometry of each 3BL's real 3D shape. Segment borders were typically where ethmoidal cells meet. We followed a vertical order from superior to inferior. We distinguished 295 segments of 100 3BLs, categorized into 9 types (Fig. 2).
The frontal plate (Fig. 2, 4th row) is basically a 2D plate in the frontal plane. The frontal wave (Fig. 2, 5th row) is similar but bears anterior and posterior protrusions and appears on axial images as a waveform. The oblique plate (Fig. 2, 3rd row) is basically a 2D plate which angles with the frontal plane. Frontal segments can be easily distinguished from obliques on axial images. The oblique wave (Fig. 2, 2nd row) is like the frontal wave but its main axis is oblique. The sigma plate (Fig. 2, 1st row) shows a sigmoid ("S"-like) angulation. The horizontal plate (Fig. 2, 8th row) is a 2D plate nearly in the horizontal plane. (We apply this nomenclature for didactical reasons.) Concave (Fig. 2, 6th row) and convex plates (Fig. 2, 7th row) develop due to ethmoidal cell bulging and refer to their shape as observed from anterior. Their combination results in a complex concave/convex plate, which will be specified later.
Furthermore, intrinsic parameters of the 3BL models (e.g. surface size) have been defined. We also studied the neighboring anatomical features of the 3BL: the bordering structures; the vertical extension of the part in the ethmoidal fovea (the distance between the horizontal level of the most superior point of the MT where it undergoes into the 3BL and the most superior point of the 3BL); the presence of pneumatized middle turbinate as well as the adherence of the ethmoidal bulla and the 3BL.
Statistical analysis. The statistical analyses have been performed using the R statistical programming language 13 (version 3.6.0, https:// www.R-proje ct. org) with the significance level set at p < 0.05 and confidence Ethics. All methods and investigations were carried out in accordance with relevant guidelines and regulations. Regulations on the tissue handling of the Dept. of Anatomy, Semmelweis University were accepted by the Senate of the Semmelweis University ("Rector act. 1. R/1/2017 VII.24"). A permission to use the samples was obtained from the institutional responsible of body donation. The CTs were anonymized and performed for other (dental) diagnostics. These belong to the database of the Dept. of Anatomy, Semmelweis University. A permission to use the anonymized images was obtained from the institutional responsible. This study is in accordance with the ethical standards of the institutional and/or national research committee, the regulating law ("1997. year CLIV. law 16. Section 2nd paragraph") and the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Based on these standards and regulations, for this retrospective study an ethical approval is not necessary. For this type of study, formal consent is not required.

Results
Geometry of the 3BL. The 3BL-a thin bone plate with occasional circumscribed thickenings up to 3 mm, covered by a mucous membrane-has been defined in each case (Fig. 3 In general, we could not observe a stringent "L"-shaped 3BL appearance with 2 portions in a lateral view. Our data showed a great diversity of 1-5 different segments (Fig. 4). On the left side the 3BL was built of 2.96 ± 0.67 segments (median: 3; range 2-4); on the right there were 2.94 ± 0.74 (median 3; range 1-5). Overall, there were usually 3 segments without significant side difference (Fisher exact test, p = 0.92). We identified a 2nd segment in 99%, a 3rd in 76%, a 4th in 19% and a 5th in 1% of all sides (Table 2). Therefore, the division of the 3BL to segments is more eligible than to an anterior frontal and a posterior horizontal part. Pearson correlation test showed no association between the number of segments and the surface area neither on the left (R = 0.09, p = 0.51, CI − 0.19 to 0.36) nor on the right side (R = 0.25, p = 0.08, CI − 0.03 to 0.49). www.nature.com/scientificreports/ In regard to their prevalence, we separated 3 main segment groups. A cohort of 3 types (horizontal, concave/convex, sigma) was present in 66 or more sides. The horizontal plate was the most frequent (68%). In the residual cases (32%) we observed a vertically oriented (not "L"-shaped) 3BL without a horizontal portion where the posterior MT attached directly to the ethmoidal crest. Examples of the second group (concave, convex) were observed on 30-33 sides. Rare types (10 or less sides) formed the third cohort.
Counting from superior to inferior, the sigma plate was the 1st segment in 55 sides, the 2nd in 11 sides but was not found further inferiorly. The concave/convex was mostly the 2nd segment (49 sides). The horizontal plate was mainly the 3rd (44 sides) or 4th (18 sides), but never the 1st segment, implying to its typical posteroinferior location.
The most frequent phenotype of the 3BL. The 9 segment types may result in an endless variability of the 3BL. Table 2 already showed the strong presence of a possible 3BL which should be built of the sigma (as the 1st), concave/convex (2nd) and the horizontal segments (3rd).
There are anatomical conditions which are associated with the sigma and concave/convex segments' development and explain their high prevalence. The sigma plate originates from the posterior expansion of the lateral anterior ethmoidal cells and/or the anterior expansion of the medial posterior ethmoidal cells. It was always anteromedial-posterolateral oriented in our material. The number of participating cells was not constant.
From an anterior point of view, the concave/convex segment appears to be convex on the sagittal images but concave on the axial images, as if it would have been shaped by a "bent tube" object, e.g. the ethmoidal bulla www.nature.com/scientificreports/ (Fig. 6). Accompanying factors were constantly observed in the presence of this segment type. Firstly, on axial images, compared to the anterior ethmoid, the posterior ethmoid was wider. Secondly, the attachment of the 3BL to the orbital lamina of the ethmoid bone was approximately 90°. Thirdly, on axial analyses, the medial border of the 3BL was located slightly anteriorly compared to its lateral border. The middle part of the 3BL is posteriorly. Lastly, a posterior ethmoidal cell (or infraorbital Haller-cell, as seen on 2 sides) produced laterally an anterior shift of the 3BL. Nota bene: the first two statements exist together.
Topographic anatomy of the 3BL. The 3BL is bordered laterally by the orbital lamina of the ethmoid bone, the maxillary bone and where appropriate by the perpendicular plate of the palatine bone (Fig. 1, Video 2,  Pneumatized MT (concha bullosa) led to deformations and increased surface of the 3BL. The average surface area of the 3BLs was 380.14 ± 111.99 mm 2 (189.80-824.22) in cases without and 443.41 ± 149.44 mm 2 (218. 20-835.21) in cases with concha bullosa. The difference between these groups was statistically significant using Mann-Whitney U test (p = 0.04, CI 1.17-119). Pneumatization originating from the anterior ethmoid was seen in these cases bilaterally in 14%, unilaterally on the left side in 4%, on the right in 5%. Pneumatization from the posterior ethmoid was seldom (2%, 3%, 0%, respectively). Comparing the prevalence of pneumatized MT on the two sides using Fisher exact test, the difference was statistically not significant (p = 0.45).

Discussion
Stammberger et al. 7,14 described the 3BL as an "L"-shaped bone plate in a lateral view with an anterior frontal and a posterior horizontal part, connecting the MT to the skull base and thus stabilizing it. As a result of the ethmoidal cells' expansion and distortion, each 3BL is unique 10 . Consequently, the intraoperative identification of this major landmark structure can be challenging despite thorough studies of CTs prior surgery. Recently the senior author provided a comprehensive overview of the ethmoidal anatomy 15 . However, there are only a few morphological analyses about the 3BL.
According to Kim et al. 8 , the 3BL might have the most uniform shape among the basal lamellas. They distinguished "the anterior portion […] in an almost frontal" (vertical) and "the posterior portion […] in an almost horizontal plane", similarly to the already mentioned "L"-shape. In their cadaveric-radiological study, the anterior portion attached to the skull base on a straight (91%) or anteriorly indented (9%) line. They categorized the relationship between the 3BL and the ethmoidal bulla whether the posterior bulla wall is the lower half of the anterior vertical portion of the 3BL (38%), or the posterior bulla wall builds the entire anterior portion of the 3BL (21%). Similar to them, we frequently (87%) found an adhesion.
Based on the bilateral analysis of 100 CTs, Wu et al. 16 observed an oblique posterior plane with circa 127° angle to the anterior part. They described the horizontal portion as a "slope". On our sophisticated 3D models, one could only broadly assess this angulation. They separated linear (22%), curved (56%) and angular (20%) appearances of the 3BL as observed on axial images. These are of great clinical value but remain artificially chosen data extracts which only partly imply to the real shape.
A shortcoming of the studies mentioned above is the subjective, schematizing and simplified description of the 3BL's intricate and diverse morphology. In this analysis we aimed to demonstrate the 3BL's 3D anatomy. We analyzed 100 basal lamellas and identified phenotypes built up from 1 to 5 pieces of 9 distinctive segment types. Our results confirm the general diversity of the 3BL. However, the high occurrence of the sigma-concave/ convex-horizontal 3BL (27%) and the sigma-concave/convex pattern (42%) may imply to a typical phenotype which we believe to be a novel information. Contrary to most reports, we could not recognize the "L"-shape of the 3BL in a lateral view. In accordance with many authors, we identified a basically vertical anterior part of the 3BL. However, after that we observed numerous segmentation patterns here. The horizontal portion lacked in 32%. Consequently, the present study showed that the commonly accepted schematic shape of the 3BL is not generally valid. This made a new classification necessary.
Diverse methods and software solutions were considered; only CBCTs (or high resolution CTs) with the 3DSlicer application enabled precise, reproducible and standardized evaluation. We sought for resemblance to www.nature.com/scientificreports/ geometrical forms which is certainly no absolute identity. It has to be noted that we focused on the appearance of the 3BL from an anterior point of view as this is the common direction of any surgical approach. Nevertheless, revisions and analysis of accompanying anatomical factors support our findings. We are not aware of any other 3DSlicer-based surgically oriented studies of the middle turbinate's anatomy. However, diverse workgroups already reported on this software's feasibility to analyze the paranasal sinuses and the skull base for anatomical 17,18 , biophysical 19,20 or forensic 21 purposes. Raappana et al. described a 3DSlicerbased planning workflow for transnasal pituitary surgery, focusing on the relevant parasellar anatomy 22 .
Dissection within the ethmoidal labyrinth should be performed with an inferomedially accented orientation to avoid unintended exposure of the anterior cranial fossa superiorly and the orbit laterally. For this reason, the segments can be reliable surgical landmark structures, especially the sigma, the concave/convex and horizontal types which showed a high prevalence in our data. The description of the variations with a focus on the sigma and concave/convex segments may be of particular clinical significance as they belong to the anterior vertical part of the 3BL which is the typical surgical dissection site. Both examples have a well-recognizable phenotype. The sigma plate was always anteromedial-posterolateral oriented and appeared to be the 1st or 2nd segment from superior (55% and 11%, respectively). Consequently, dissection of the sigma plate likely leads to the superior part of the posterior ethmoid. The concave/convex was the 2nd segment in 49% but the 1st in 9%. According to our data, dissection of the concave/convex segment likely leads to the inferior part of the posterior ethmoid. Thus, the concave/convex segment's peculiar shape may be an additional information besides other well-known anatomical landmarks to define the pathway to the posterior ethmoid.
The triplanar analysis of axial, sagittal and frontal 2D reconstructions is an efficient method and routinely applied in the daily care. Commonly, the 3BL is understood as a continuous structure with a variety of curvatures without definitive angles, which could be a criticism to the statements of the present work. Our study is based on this widely accepted concept, but additionally represents how exactly the 3BL's "fluid" shape can be derived from the accompanying ethmoidal cells and which factors influence the 3BL's shape. We hope that this descriptive manuscript provides a better understanding of this delicate anatomical area for morphologists and surgeons. We believe that the 3D depiction of the basal lamella enables not only to understand its true shape but to accurately interpret the triplanar 2D reconstructions and assess the lamella's and the neighboring structures' topography. CT analysis prior to surgery can help to identify the desired entry zone and separate it from the superior and horizontal portions of the 3BL. The segmenting method is eligible, as it relies on the simple observation that the borders of segments are also the borders of ethmoidal cells. We identified a 3BL-phenotype with notably high prevalence and strong relation to given accompanying factors, serving as the major observation of this study. We suggest to choose a surgical approach after the geometrical details of the ethmoid were configured beforehand. However, the clinical impact of our findings has to be confirmed by surgeons and needs further research.

Conclusion
The 3BL is an important landmark for endoscopic sinus and skull base surgery with variable morphology. Studies using 3D reconstructions gained from CT scans allow a precise description of the 3BL. We introduced a segmental anatomical description which is more applicable than dividing them into an anterior frontal and a posterior horizontal part. The sigma-concave/convex-horizontal trisegmental and the sigma-concave/convex bisegmental pattern were common suggesting their close relation to the typical 3BL phenotype. This study can help surgeons to predict anatomical variations during endoscopic sinus and skull base procedures and improve their intraoperative orientation during ethmoidal dissection.