Evaluating micro-computed tomography for investigation of the pediatric hyoid-larynx complex

Five pediatric hyoid-larynx complex samples, ranging from 12 days to 16 years of age, were obtained from forensic autopsies (February–August 2022). Based on the preliminary information provided at the time of case referral, suspected neck trauma was considered unlikely in three cases, deemed possible in one case, and was suspected in another case (Table 1).

During the autopsy, the on-duty forensic pathologist (including V.S.M.) excised the hyoid-larynx complex in accordance with local standardized procedures. This process required an incision in the neck, followed by the removal of the superficial throat muscles and thyroid gland. The hyoid-larynx complex was then carefully dissected, along with the tongue and the upper section of the trachea. Once isolated, the tongue was separated, allowing for a manual assessment of the stability of the hyoid bone and superior thyroid horns. Each excised hyoid-larynx complex, together with the surrounding soft tissue if present, was subsequently preserved in 4% formaldehyde.

Before imaging, the samples were thoroughly rinsed with phosphate-buffered saline (PBS) solution and placed in small padded plastic containers to ensure stability during scanning.

In order to perform diceCT, the samples were stained. Traditional Lugol’s solution is acidic and can cause tissue shrinkage [18]. To counter this effect, a buffered variant, Buffered Lugol’s solution (B-Lugol), has been developed in our lab to reduce tissue shrinkage while preserving staining effectiveness [18]. Contrast enhancement was achieved by submerging the specimens in 3.75% B-Lugol, using a solution volume equivalent to 20 times the weight of the sample (w/v). Based on our previous experiences and the hyoid-larynx complex’s weight, the staining process lasted between 4 and 16 days (Table 2).

Micro-CT and diceCT imaging were performed by a trained micro-CT technician (D.D.) using the Phoenix Nanotom M (GE Sensing & Inspection Technologies GmbH, Pforzheim, Germany). For the largest sample, a different micro-CT system, the TESCAN UniTOM XL (TESCAN, Brno, Czech Republic), which is equipped with a higher-powered radiation source, was used. The samples were scanned at voxel sizes between 12 µm and 35 µm. A detailed overview of the scan parameters is presented in Table 3.

Standard forensic assessment for pediatric cases includes total-body CT, which was performed in all cases prior to the autopsy. The total-body CT imaging was done using a clinical CT scanner, Revolution CT (GE Healthcare, Chicago, IL) and included a protocol with voxel sizes of 625 µm. The scans were assessed by one of two clinical radiologists with experience in forensic imaging. Scans and reports were retrieved to compare with the micro-CT and diceCT scan results.

Images were processed using Amira Software version 3D 2021.2, which was also used for visualization and 3D volume rendering (Thermo Fisher Scientific, Waltham, MA). Anatomical variations, ossification stages, potential fractures or hemorrhages, and notable hyper- or hypodensities were assessed by a trained analyst (G.M.M.T.) and two experienced forensic (pediatric) radiologists (H.M.D.B. and R.R.V.R.). These assessments were performed visually. Because micro-CT data are presented in arbitrary grayscale units rather than standardized Hounsfield Units (HU), interpretation relied on relative contrast differences between tissues. This allowed differentiation between mineralized and non-mineralized components, as well as detection of subtle density variations within cartilage.

All samples exhibited incomplete ossification on micro-CT imaging (Fig. 3). The youngest sample (12 days old) showed ossification limited to the hyoid body, with cartilaginous greater and lesser horns of the hyoid and laryngeal structures. The other samples showed (almost) complete ossification of the hyoid body and greater horns. Only the oldest sample showed an ossified lesser horn of the hyoid on the right side and one circular ossification center in the thyroid cartilage (Fig. 4).

Bild vergrößern

Bild vergrößern

All pediatric hyoid-larynx complexes had a non-fused hyoid bone and normal greater horns (Fig. 3). Samples 1–4 contained bilateral cartilaginous lesser horns and no anomalies regarding the superior thyroid horns. The sample of the 16-year-old (case 5) exhibited a right-side ossified lesser horn, a left-side cartilaginous lesser horn, and bilateral triticeal cartilages. The sample of the 14-year-old (case 4) contained two bone fragments on the stylohyoid ligament. Additionally, samples 3 and 5 showed small circular defects in the lamina of the thyroid cartilage (Fig. 4).

Complete staining with B-Lugol was achieved in three specimens. These samples weighed 4–17 g and were completely stained in 4–5 days (Table 2). The largest pediatric sample weighed 121 g, and only the outer layers of the hyoid-larynx complex were stained after a staining time of 16 days. This resulted in stained bone and cartilage tissue, while the soft tissue in the center of the sample remained unstained.

DiceCT revealed uni- or bilateral hyperdense bundles crossing through the thyroid lamina in samples 3 and 5 (Fig. 4). A hyperdense ossification center was identified within the thyroid cartilage in the oldest pediatric sample and in addition some hypodense aspects were found in the cricoid cartilage (Fig. 4). In all specimens, a line with higher density was noted encircling the hypodense cartilage parts of the hyoid (if not completely ossified) and larynx on diceCT (Figs. 4 and 5). Figure 5 shows a 3D diceCT volume rendering and anatomically marked transverse sections of sample 1 to show the detailed visualization of the pediatric hyoid-larynx complex’s anatomy.

Bild vergrößern

The total-body CT and autopsy reports revealed no traumatic findings. On micro-CT and diceCT, no fractures or hemorrhages were identified in any of the pediatric samples, which is consistent with the total-body CT and autopsy findings.

All pediatric specimens demonstrated a non-fused ossified hyoid bone with either bilateral or unilateral cartilaginous lesser horns (Fig. 3). These findings align with known ossification patterns of the hyoid and laryngeal structures, as hyoid ossification starts within the first two living years and calcification of the larynx does not start before the second decade [2]. In addition, literature reports that the calcification of the larynx starts within the inferior thyroid horns and spreads to the superior thyroid horns [2, 4]. Nevertheless, our results show a sample of a 16-year-old male that already showed one calcified center in the thyroid lamina as the beginning of ossification (Fig. 4). This center was identified as an ossification center as Claassen et al. [20] found that the laryngeal ossification pattern consists of a special mode of endochondral ossification, which includes the mineralization of cartilage and creating islands of cartilage before being covered by the deposition of bone. The hyperdense center could represent the mineralized cartilage, as this cartilage appears denser than bone on X-ray images [20].

Furthermore, hypodense features were identified within the cartilages of the hyoid-larynx complex of this 16-year-old (Fig. 4). These features could present part of the ossification process. Dedivitis et al. [21] studied histological aging changes in the cricoid cartilage and showed that adolescents’ cricoid cartilages only consist of typical hyaline cartilage, while the older cartilages showed central areas of lamellar bone tissue and bone marrow cavities filled with adipose and/or hematopoietic tissue. These cavities could possibly be seen as hypodense irregularities in the larynx cartilages on micro-CT.

In addition, a denser layer was observed around the hypodense cartilages in all pediatric samples (Fig. 4). This layer appeared as a hypodense line surrounding the even more hypodense laryngeal cartilages, and in the younger samples, it also encircled the hypodense cartilaginous greater horns of the hyoid (Fig. 5). This could shed light on the cartilage and bone development, as it was not observed in adults [14]. The dense layer could present the outer layer of hyaline cartilage, the perichondrium. With age, the cartilage is subject to the process of endochondral ossification, resulting in the perichondrium being converted into periosteum [20]. Periosteum is a thin layer of membranous connective tissue, while perichondrium is denser and consists of fibrous tissue [22]. Besides, the perichondrium appears thinner on histological slices of the cartilaginous adult larynx in comparison with the pediatric larynx; therefore, it could show less dense on diceCT as well [19].

In two samples, a uni- or bilateral circular defect in the lamina of the thyroid cartilage was observed on micro-CT (Fig. 4). In the literature, this opening is known as the foramen thyroideum (FT) [23]. The FT can occur in one or both laminae of the thyroid cartilage. We found a FT in two of the hyoid-larynx complexes (40%), in one sample unilateral and in the other one bilateral. Previous research reports an incidence of FT in adults of 24–31%, of which 2–11% were found bilaterally [23‐25]. The prevalences found in this study showed a slight difference in comparison to current literature, which could be due to our small sample size. According to León et al. [25], the adult FT contains a neurovascular bundle in 73%, a nerve branch in 20%, and in 7% only a vascular branch. In our research, diceCT showed a hyperdense bundle within every FT. Within those hyperdense bundles, two hypodense tubular structures could be visualized in one sample, and the other sample showed one hypodense and one hyperdense tubular structure (Fig. 4). Since blood shows hyperdense on diceCT images, we suspect that the hypodense tubular structures are nerves (branch of the superior laryngeal nerve) and the hyperdense structures are blood-filled vascular branches. As neurovascular bundles are surrounded by connective tissue, the hyperdense staining surrounding the bundles is considered connective tissue. This is in line with the hyperdensely stained tissue we discovered in the cricoid area, paraglottic, and pre-epiglottic spaces (Fig. 5). These spaces contain areolar or loose connective tissue, which is composed of adipose tissue with elastic and collagen fibers [22]. Hence, B-Lugol binds to adipose tissue and/or elastic and collagen fibers as well. Anatomical dissection could provide a definitive answer to the content of the FT.

The observed ossification process, hyperdensely stained perichondrium, and FT provide potential new insights into early hyoid-larynx complex development and anatomy, highlighting the potential of micro-CT and diceCT for developmental anatomy research.

The forensic relevance of these imaging techniques lies in their ability to non-destructively assess the pediatric hyoid-larynx complex in micrometer-level detail, something previously unattainable using standard imaging or autopsy methods. Partial ossifications and transitional zones, which are difficult or impossible to assess macroscopically, become clearly visible. In contrast to traditional histology, diceCT enables three-dimensional mapping of hemorrhages in situ and in anatomical context, without requiring destructive tissue dissection [20]. This allows for more precise, targeted histological sampling and preserves the integrity of delicate pediatric structures. Second, the high resolution of micro-CT and diceCT makes it possible to identify subtle trauma and visualize hematomas with confidence [20]. Given this level of detail, a negative scan may reasonably exclude significant hyoid-larynx trauma. With advancing technology and growing interest, micro-CT and diceCT may in time become valuable tools for routine forensic evaluation. Importantly, current forensic protocols do not recommend separate imaging of the excised pediatric larynx and hyoid, largely due to the insufficient resolution of conventional modalities. Now that micro-CT offers a viable alternative, there is a strong case for systematic research into its forensic value, with the potential to revise and improve existing guidelines. Third, the ability to clearly visualize ossification centers and perichondral layers could support forensic age estimation. Finally, the resulting 3D volume renderings are not only scientifically informative but also provide clear and compelling visualizations for courtroom presentations. Compared to traditional autopsy photos or histological slides, these high-resolution images enhance communication of complex anatomical findings to legal professionals and lay audiences alike.

This study’s small sample size limits the generalizability of the findings. Additionally, incomplete staining in the largest pediatric sample highlights the need for optimized staining protocols. In hindsight, this sample contained a lot of soft tissue and, based on its weight compared to the other samples, required a longer staining time. Based on our experience, we estimate that this particular sample would have required approximately 30 days of staining to achieve complete and uniform contrast enhancement. To strengthen the forensic applicability of diceCT, future research should include trauma cases and incorporate histological validation to confirm the accuracy of imaging findings. Finally, some practical limitations of micro-CT should be acknowledged: the technique requires access to high-end imaging infrastructure, is associated with significant costs, and demands specialized expertise for both image acquisition and interpretation. These factors may currently limit its widespread adoption in routine forensic workflows, particularly outside academic or research-oriented settings.