Evaluation of a multi-material 3D printed temporal bone as a training tool: the Australian otolaryngology trainee experience

Introduction

Temporal bone anatomy is a complex study that Otolaryngology registrars are required to master by the end of their training. Dissecting a cadaveric temporal bone is becoming an increasingly rare opportunity for most trainees due to limitations of access to cadaveric dissection (costs associated with cadaver laboratory access, ethical considerations, and limited donations of human bodies). Access to three-dimensional (3D)-printed models may improve the opportunities to understand the anatomy. The Australian Society of Head and Neck Surgeons (ASOHNS) training program requires trainees to have completed sixty temporal bone dissection exercises by the completion of training (1). Various 3D-printed models have been analysed for usefulness as a training adjunct for temporal bone dissection due to increasing scarcity of training registrars’ access to cadaveric dissection (2-19). Biomedical laboratories with access to 3D printing capabilities and expertise allow for innovation in this setting of demand (13). The Herston Biofabrication Institute (Metro North Health Service, Queensland) developed 3D-printed temporal bone models using open-access micro-CT data (12-micrometer slices) with auto-segmentation for some anatomical parts. Blender software was used to model intricate anatomical details, with the guidance of a skull base surgical fellow and the models were fabricated using a multi-material jet printer (Stratasys J750), which produces assembled materials of varying hardness qualities. For further details on the development of the temporal bone model, please refer to Kanaganayagam et al. (20).

Chien et al. (1) established, via semi-structured questionnaires using a Likert scale completed by Otolaryngology trainees, a 3D printed temporal bone model as a validated tool for developing skills in temporal bone dissection in general and for specific procedures. Participants were enthusiastic about the use of the models in training courses. McMillen et al. (3) compared various materials and printing methods and analysed the feedback of consultants and trainees who dissected their models and found resin models to perform superiorly. Mowry et al. (21) compared various printing methods and usefulness in specific anatomical procedures with stereolithography and resin based models performing superiorly. Frithioff et al. (12) identified the usefulness of 3D printed models in training and the transferrable skills and development of surgical skills compared to virtual reality training.

This study aims to evaluate the anatomical fidelity and practical utility of a multi-material 3D-printed temporal bone model compared to cadaveric specimens for otolaryngology trainee education. Additionally, the study explores the model’s usefulness for trainees at different levels of surgical expertise. It was hypothesized that the model’s utility and the ability to complete dissection tasks would correlate with the level of surgical training.

Methods

The model as a training tool

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Gold Coast University Hospital Human Research Ethics Committee (HREC/2022/QGC/87159). Informed consent was taken from all individual participants. The study is reported according to the SURGE reporting guidelines (available at https://www.theajo.com/article/view/10.21037/ajo-24-59/rc). A temporal bone model was designed with open-access data from a cadaver which provided high quality and high-resolution slices of a micro-CT (12 micrometer resolution) of a cadaveric temporal bone. This was beneficial as the most important anatomical parts were auto-segmented parts. This included the facial nerve with the chorda tympani, the sigmoid sinus, the carotid artery, the tympanic membrane and the ossicles (4). Blender software was used to create parts that required further detail in order to make the anatomy more realistic with the guidance of a surgical registrar and skull base surgical fellow. This process was required for the following structures that were not already present in the Delta dataset: (I) the internal acoustic nerves (facial nerve, cochlear nerve, superior and inferior vestibular nerves) with a dural sheath, (II) the geniculate ganglion in continuity with the facial nerve, (III) first and second genu of the facial nerve in continuity and a facial nerve exiting the stylomastoid foramen, (IV) a prominent digastric ridge modelled into the mastoid bone, (V) a prominent stylomastoid process for readily available identification, (VI) incudo-malleolar joint and incudo-stapedial joints modelled as separate structures so as to be printed with a separate material from the bone to allow the ossicles to be mobile, (VII) the stapedial tendon and ossicular ligaments to allow suspension of the middle ear bones within its space, (VIII) a superior semicircular canal projected to the middle cranial fossa to resemble a prominent arcuate eminence, (IX) modelling of the otic capsule with perilymph and endolymphatic systems, (X) communication of the mastoid air cells a with drainage hole to allow flushing of filler material from the scaffolding of the model, (XI) sinus dura mater (20). The data is then assembled in a 3D printing software (GrabCAD) and this information is translated to the 3D printer with the various anatomical parts requiring varying material properties (hardness, softness and colour). The Stratasys J750 printer was selected due to its unique capabilities of printing different anatomical parts assembled together while maintaining their unique anatomical properties (for example, a soft coloured facial nerve within a hard temporal bone) without having to add these soft parts into a model after printing.

Temporal bone dissection and evaluation

This study was conducted in March 2023 at the Queensland Temporal Bone course in Brisbane during the Asia-Oceania Otolaryngology Head and Neck Surgery Congress. The participants included registrars from the Australian Society of Otolaryngology Head and Neck Surgeons (ASOHNS) training program who after completing a cadaveric dissection, were invited to participate in a questionnaire after dissecting the 3D printed temporal bone model. Each trainee received a 3D printed model that was not previously dissected. Trainee levels (novice, intermediate, competent) were recorded. Written consent was obtained from all trainees prior to their participation in the study. No incentives or inducements were offered to encourage participation.

The participants performed otologic procedures: cortical mastoidectomy, posterior tympanotomy, raising a tympanomeatal flap, identifying middle ear anatomy and dividing ossicles (as seen in Figure 1), identifying the basal turn of cochlea, performing a labyrinthectomy and dividing the components of the internal acoustic canal. They then completed a de-identified hard copy survey questionnaire where they answered questions on their ability to perform the otological procedures on the models. Participants provided feedback on the performance of the model for the activities that they learned in the course, as well as provided feedback on the usefulness as a training tool. Participants were only contacted at the time of the temporal bone course, and did not provide any further feedback with follow up contact. Participants were approached by the first author by email with a consent form prior to the course and in person at the temporal bone course. Participants were provided with a 3D printed temporal bone, a temporal bone holder, dissecting equipment at the temporal bone dissection course, consent forms and a questionnaire.

Figure 1 Cadaveric ossicles compared to three-dimensional printed model ossicles excised in the middle ear exercise.

A questionnaire was designed to acquire feedback from training Otolaryngology registrars on the performance of the model in surgical procedures, and quality of landmarks of a temporal bone dissection (Appendix 1, Questionnaire A), followed by questions on the quality of the 3D Printed Temporal bone model as a training tool (Appendix 1, Questionnaire B). This questionnaire was adapted from the validated tool by Chien et al., with consultation with a senior neuro-otologist (the final author). The first part of the questionnaire (Appendix 1, Questionnaire A) followed a sequence of questions which allowed the participants to dissect the models as they would a cadaver and provide feedback on the ability of each individual step of the dissection tasks on the model. The questionnaire asked 20 questions which were grouped into: (I) the participant’s level of training (a variable factor that may influence the ability to complete the procedures), (II) the anatomical and realistic performance of the model in the procedures, (III) the usefulness of the model as a training tool, and (IV) the realism of the set up. The answers were scored on a five-point Likert scale. An answer of ‘four’ represented that all of the landmarks were identified, and all manoeuvres were achievable at a level comparable to cadaveric dissection. ‘Three’ represented that most landmarks were identified and most of the manoeuvres achievable when compared to the cadaveric dissection. ‘Two’, that some landmarks were identified and some of the manoeuvres achievable when compared to the cadaveric dissection. ‘One’ represented that the landmarks were not identified, and the manoeuvres were not able to be performed when compared to the cadaveric dissection (N/A), which was for manoeuvres that were not achieved and was scored as ‘zero’. No answer was also scored as ‘zero’.

A score of ‘three’ and above indicated that the 3D-model was a comparable tool for temporal bone dissection. A score of ‘two’ indicated that the model was able to perform as a tool but not at a reasonable training standard. A score of ‘one’ and below indicated that the model was not suitable in likeness to a cadaver to complete the step of dissection.

The second part of the questionnaire (Appendix 1, Questionnaire B) focused on the participant’s score of usefulness of the model as a training tool and was scored from 1–5: strongly disagree [1], disagree [2], neutral [3], agree [4] and strongly agree [5].

For the purpose of this study, a complete questionnaire was defined as one in which all sections, including demographic information and all Likert-scale evaluations, were fully answered. A partial completion was defined as a questionnaire where one or more sections were incomplete, but sufficient data were provided to analyse at least one aspect of the educational utility of the specimens. Only complete responses were included in the primary analysis, while partial completions were analysed separately to assess potential response patterns or biases.

Participants’ level in training

Participants also answered which stage of the ASOHNS’ three levels of training (‘Novice’, ‘Intermediate’ and ‘Competent’) they were currently at when performing the dissection course. Novice registrars are typically within their first two years of surgical training. Intermediate registrars are within their third and fourth years of training, and competent registrars are in their final year of training. This question was asked to as focus on the usefulness of the model for different stages in training.

Statistical analysis

Descriptive statistics, including mean Likert scores with standard deviations, were used to analyse the questionnaire data. The data were further stratified by trainee level for subgroup analysis.

Results

Fourteen out of eighteen trainee registrars at the course participated in the study (78% participation). Of those who participated, 11 were novice level, 2 were intermediate level, and 1 was a competent level trainee. At the time of the study, there were seventy-eight full time registrars in training. The participants’ representation captures 18% of the population of Australian trainees at the time, in the study. Of the enrolled trainees at the time of the course, 23% attended the temporal bone course. See Table 1 for results and Appendix 1, Questionnaire A for the Likert questionnaire in full.

The mean Likert score for the ability to perform a cortical mastoidectomy, identify the sigmoid sinus and sinodural angle was 3.71 [standard deviation (SD) =0.47, response rate 100%]. The mean Likert score for the ability to expose the mastoid portion of the facial nerve using anatomical landmarks, including the short process of the incus and the lateral semicircular canal was 3.64 (SD =0.63, response rate 100%). The mean Likert score for the ability to perform a posterior tympanotomy was 2.64 (SD =1.82, response rate 71%). The mean Likert score for the ability to identify the components of the ossicular chain and dissect the individual components was 2.36 (SD =1.69, response rate 71%). The mean Likert score for the ability to visualise the basal turn of the cochlea and the round window niche was 1.42 (SD =1.79, response rate 43%). The mean Likert score for the ability to identify the petrous internal carotid artery and the Eustachian tube opening was 1.07 (SD =1.77, response rate 29%). The mean Likert score for the ability to perform a labyrinthectomy and identify the posterior, superior and lateral semi-circular canals and the vestibule was 1.76 (SD =1.89, response rate 50%). The mean Likert score for the ability to skeletonise the internal auditory canal (IAC) and open the canal was 0.571 (SD =1.45, response rate 14%); however, this was only answered by two participants and was answered by the rest of the participants as “not applicable”. For those who did answer (one was in novice and one was in intermediate years), both scored 4 on the Likert scale. Similarly, the mean Likert score for the ability to identify the facial nerve within the IAC was 0.36 (SD =1.08, response rate 14%) and was only answered by the same two participants. The scores for the ability to identify the cochlear nerve, superior and inferior vestibular nerves within the IAC were also low with a mean score of 0.71 (SD =0.27, response rate 7%), 0.71 (SD =0.27, response rate 7%), and 0.71 (SD =0.27, response rate 7%) respectively with only one participant answering these questions. The rest of the participants answered “not applicable”. In the free feedback questionnaire section, participants stated they did not have enough time to complete these or were not able to complete these procedures.

Figure 2 demonstrated the participants’ Likert score responses to the dissection and reflect the model’s high performance in the surgical procedures except for the internal acoustic canal surgery (tasks 9, 10, 11, and 12), which was only completed by two participants and scored N/A by all other participants.

Figure 2 The mean score given by 14 participants for each surgical procedure evaluated in the three-dimensional printed temporal bone model. Complete question descriptions are provided in Appendix 1, Questionnaire A.

Figure 3 demonstrates the responses of participants to Questionnaire A, question 1. This shows excellent performance of the model in the cortical mastoidectomy procedure, including identifying the sigmoid sinus, sinodural angle and tegmen across all participants; novice, intermediate and competent with a median score of 4.

Figure 3 The Likert score given by each of the 14 participants on the model’s comparability to perform a cortical mastoidectomy, identify the sigmoid sinus and the sinodural angle.

Figure 4 demonstrates the responses of participants to Questionnaire A, question 7. This shows the performance of the model in a labyrinthectomy which was performed by six novice participants and one intermediate participant who scored between ‘three’ and ‘four’. Although only seven participants completed this procedure, those who completed the step rated the model as comparable to a cadaveric dissection.

Figure 4 Likert scores for labyrinthectomy performance per participant.

The second part of the questionnaire (Questionnaire B) focused on the participant’s score of usefulness of the model as a training tool and was scored from 1–5: strongly disagree [1], disagree [2], neutral [3], agree [4] and strongly agree [5].

In terms of the model setup being realistic for training, it achieved a mean score of 3.85 (SD =0.66, response rate 100%). It scored 2.2 (SD =1.50, response rate 100%) in whether a distracting odour was noticed, indicating that most trainees did not detect an odour during dissection. A mean score of 4.21 (SD =0.97, response rate 100%) was achieved on being useful for skill development based on the participants’ level of training. It scored a mean of 3.80 (SD =0.80, response rate 100%) on being able to practice new skills. In answering the question about whether the dissection could be applied to their surgical practice, a mean score of 4.29 (SD =0.72, response rate 100%). Those who strongly agreed were Novice trainees.

In answering whether the model would be useful to practice dissection and improve their procedural skills in their own time, participants scored 4.14 (SD =0.77, response rate 100%).

In answering whether the model would be helpful for training advanced Otolaryngology, Head and Neck surgery trainees, the mean score was 3.78 (SD 1.18, response rate 100%). The two participants who did not agree that it would be useful were at intermediate and competent level and may reflect diminishing returns on using this as a training tool with increasing trainee experience.

In answering whether the participants would like to pursue a career in otology, neuro-otology or skull base surgery, a score of 3.35 (SD =1.15, response rate 100%) was given reflecting a varied response in interest in the subspecialty.

Novice registrars

There were eleven novice registrars who participated in this study. All novice participants were able to complete the cortical mastoidectomy and identify the sigmoid sinus, facial nerve and tegmen (mean score 3.72). They all identified the mastoid portion of the facial nerve and landmarks of the lateral semicircular canal and long process of the incus (mean score 3.64). In performing a posterior tympanotomy, eight participants completed this (mean score 3.63). Six novice registrars were able to identify the components of the ossicular chain and dissect the individual components (mean score 3.25). Five were able to identify the basal turn of the cochlea and round window niche (mean score 3.2). Three were able to identify the petrous carotid artery and Eustachian tube (mean score 3.6). Six were able to complete a labyrinthectomy (mean score 3.5). One out of eleven of the novice registrars was able to identify the internal acoustic canal and divide the facial nerve within (score 4); however, was not able to complete this activity or identify the other nerves within.

In the free-text feedback, the novice participants reported that they would have benefited from having more time to complete the procedures. Two reported that they wanted more colour differentiation between the dura and the bone. Two reported that they wanted something to rest their hands on during dissection. One participant reported that the model holder was excellent and that the bone texture was excellent in hardness quality during drilling. One participant reported excellent replication of the middle ear anatomy and ossicle movement.

Intermediate and competent registrars

There were two intermediate registrars who participated in the study. Both registrars were able to perform a cortical mastoidectomy (mean Likert score 3.5), expose the mastoid portion of the facial nerve (mean Likert score 4) and perform a posterior tympanotomy (mean Likert score 3.5). One completed the dissection to identify the components of the ossicular chain (mean score 4), expose the basal turn of the cochlea (mean score 4) and identify the round window niche (mean score 4) and even proceeded to a cochleostomy. An implant was not able to be passed into the cochlea due to filling material within this part of the model. One intermediate registrar was able to identify the petrous internal carotid artery and perform a labyrinthectomy and skeletonise the IAC (mean score 4), however was not able to identify the nerves within (score 1). On overall feedback, one reported the footplate of the stapes was disarticulated on finding the incudostapedial joint and that the internal acoustic canal needed to contain nerves. They were enthusiastic on the ‘ability to perform a middle cranial fossa approach’ and found the model to be excellent overall ‘for teaching junior trainees comprehensive anatomy.

There was one competent registrar who participated in the study. The competent participant reported that they could perform the cortical mastoidectomy and expose the mastoid portion of the facial nerve and its landmarks. They did not complete any of the other exercises but reported that the setup was realistic for training purposes, and that they would find the model useful to dissect to improve procedural skills.

Discussion

In this study, we present the trainee’s perspective of dissecting the 3D printed temporal bones and evaluate the model’s usefulness in learning procedures. As hypothesized, the utility of the model varied with trainee experience, with novice trainees benefiting most from the anatomical accuracy and procedural simulation.

The model was reported to be more useful to junior training registrars. More senior trainees (Intermediate and competent levels) identified that it was not useful for their level of training but gave feedback that it would be useful to identify anatomical landmarks as a junior registrar.

Junior registrars also reported that they would find the model useful for improving their skills and could be applied to their surgical practice. However, across all levels of training, the dissection of the cortical mastoidectomy, identification of the facial nerve and examination of the ossicles within the middle ear was achievable with high mean scores (Figures 3,4).

Like Chien et al. (2) registrars found the cortical mastoidectomy, posterior tympanotomy most similar to the cadaveric dissection. This study’s scores were 1.75 (SD =1.89), however, unlike their study, the scores of the participants who were able to complete the labyrinthectomy (those who did not score “not applicable”) scored a mean Likert of 3.57 with most landmarks identified and most procedures were able to be performed comparable to a cadaveric dissection. This may account for the multi-material printer’s capabilities of representing the minute structures with the model and the removal of filler material. Chien et al. used powder deposition printing with added colour post printing to smaller parts. Similar to the Chien et al. study, registrars found that the model was an overall useful tool for training. Frithioff et al. (13) evaluated the cost effectiveness of temporal bone models made from extrusion printers (low maintenance, low cost of materials, low cost printer) with good feedback from trainees at various stages of training in a similar temporal bone course. Their results showed good feedback on anatomical landmarks of a mastoidectomy; however, they did not evaluate the other procedures required to be completed in temporal bone dissection training. In their systematic review of 36 other studies focused on 3D printed temporal bone dissection in surgical training, Frithioff et al. (22) identified the requirement of increased anatomical accuracy for more detailed parts of the procedures required in further advanced surgical training which is not necessarily required for the novice level of training in identifying surface and immediate landmarks of a temporal bone dissection which can be achieved with less expensive printing methods. They also recommended future studies on the integration of 3D printed models in training curriculum e.g., Novice trainees performing cortical dissections on these models and graduating their skill practice to the scarcely available cadavers and including virtual reality as part of the training.

The study highlights the model’s anatomical accuracy, realistic texture, and usefulness for foundational training in performing the required exercises at a temporal bone course. The results suggest that this model is particularly useful for novice trainees for anatomical landmark identification and foundational surgical skills, including visualisation of anatomical landmarks and conceptualisation of the anatomy, a realistic set up, the ability to practice using the dissection tools themselves and realistic texture and haptic feedback.

The first limitation to discuss is the selection bias introduced through the 77.8% participation rate of ASOHNs trainees attending the temporal bone course. The captured demographic also represents only 18% of the population of all trainees at the time of the course. For attendees who did not participate, their levels of training were not recorded and as such this could have created bias in the results positively or negatively (i.e., more senior trainees may have deemed the model less useful as a training tool). This could be addressed in future studies by including larger numbers of participants. The results are also skewed to a higher number of novice participants who are more likely to participate in temporal bone courses for the requirement of meeting the novice requirements of training. This however, is not a significant limitation, given that these participants are most likely to benefit from more dissection practice and therefore most likely to benefit from 3D printed temporal bone practice.

Other limitations include the use of a new questionnaire tool adapted from the Chien et al.’s study with questions that have not yet been validated. Another limitation is the lack of procedural assessments, and missing details in the internal acoustic canal anatomy. Some of the weaknesses of the model identified through survey and feedback were that the nerves within the internal acoustic canal were missing or not long enough. This was an error with the print used in this study and could be rectified for future prints. To make the model more useful to intermediate and competent trainees, improving the model with regards to detail within the internal acoustic canal and including pathology would be more appropriate for their level of training. A consideration for intermediate and competent trainees would be to use different models for specific tasks, such as division of the nerves in the internal acoustic canal, translabyrinthine approach and a middle cranial fossa approach. This study also focused on the face validity of each participant’s reporting of the exercises performed on the model which does not take into the account the participant’s ability to perform the procedure, although this has been considered in the division of participants into their levels of training. The study is also limited by the trainee’s ability to perform the various procedures and identify the anatomy. This was partially addressed by dividing the groups into ‘novice’, ‘intermediate’ and ‘competent’. The low respondent rate to some of the more difficult procedures, such as labyrinthectomy, division of the nerves within the internal acoustic canal, can be attributed to by the participant’s ability to identify anatomy and perform the procedures required in that step. This is seen in that these procedures were only answered by ‘intermediate’ and ‘competent’ participants. The responses to these questions therefore have less statistical significance.

Participants also reported difficult colour differentiation and identification of the dura. Colour differentiation can be increased by selecting various quantities of colour and hardness in the cartridges of the printer (a maximum of 3 colours on a tray) but may compromise the realism with the brightness of these (22) colours. See Figure 5 for the current colour contrast. Colour could also be added in post processing with painting but does require more labour costs to do this. Participants performed the dissection on the model after completing their exercises on the cadaver and reported a limitation in time allowed for them to perform the dissection on the model. This may be extrapolated as to why most participants reported ‘N/A’ or 0 to questions 8–12 (skeletonising the internal acoustic canal and its structures within).

Figure 5 Dissection of three-dimensional model with coloured facial nerve and sigmoid sinus.

The Stratasys J750 multi-material printer is an industrial printer and costly to purchase for each biomedical department, with each cartridge of print material approximately AUD1,100. On a run of a full tray of printing, the cost per model is approximately AUD220 of materials which does not include the labour costs for post-processing the models (cleaning support materials and removal of drainage insertion hole).

Future studies (with models with improved internal acoustic anatomy) would be helpful in assisting how these models could be integrated into training curricula and assessment of their use and potential benefits throughout the training program. This could also be incorporated into the use of different types of printing methods based on different types of procedure targets, for example, fused deposition printed models for cortical mastoidectomy and multi-material printers for acoustic canal procedures. Future studies into the cost analysis of these materials, different printing methods and models for use based on levels of training would also be useful and should focus both on how to make 3D printed models more economical, as well as the comparative cost with cadaveric temporal bone dissection.

Conclusions

The 3D-printed temporal bone model developed by the Herston Biofabrication Institute is a valuable training tool for novice otolaryngology registrars, particularly for procedures like cortical mastoidectomy and posterior tympanotomy. Almost all participants were able to perform a cortical mastoidectomy with identification of the landmarks and more than half were able to identify the middle ear structures, the basal turn of the cochlea and round window niche. Overall, the model was scored highly in usefulness as a training tool and in simulation of procedures. The feedback was that the model was an excellent tool with potential to be even more realistic, with more colour and nerve differentiation and that it would be beneficial for training and learning procedures. Further larger studies on the surgical skill acumen of trainees who practice on 3D printed models, such as a first step in dissection practice before using scarcely available cadaveric dissection or in real world practice, would be useful in determining whether these could be integrated into a training paradigm.

The model performed well in its anatomical landmark identification and simulation of dissection procedural steps. Dissection was most useful for junior otolaryngology registrars to practice dissection skills and identification of landmarks. The model set up and holder were easily replicable to a realistic dissection set up and most registrars would find the model useful to dissect in their own time outside of a temporal bone laboratory. It is likely that the model would have scored even better if participants had more time allocated to dissection.

Acknowledgments

We would like to sincerely thank Dr. Michael Redmond and Dr. Andrew Lomas for their support throughout the project, and Dr. Shane Anderson for his support during the QLD temporal bone dissection course. We also thank Mr Allan Evans of William Ross State High School QLD for providing materials for the temporal bone holders at this course.

Footnote

Reporting Checklist: The authors have completed the SURGE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/ajo-24-59/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-24-59/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-24-59/prf

Funding: This study received Regional Training Subsidy for E.K. to travel to course.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-24-59/coif). E.K., J.B., D.F., R.F. and M.W. are receiving support from Herston Biofabrication Institute (HBI, Herston Biofabrication Institute, Metro North Hospital and Health Service, Brisbane, Queensland, Australia), which is the location of corresponding author’s University of Queensland MPhil Candidate placement. E.K. reported the use of Specialist Training Program Commonwealth grant funding for travel to temporal bone course in Brisbane, from trainee’s place of work (Townsville). J.B. is a University of Queensland (UQ) Supervisor for UQ Masters of Philosophy Candidate. D.F. is a Herston Biofabrication Institute Senior Biomedical Engineer and Industrial Designer, and an Adjunct Research Fellow of the University of Queensland. R.F. is an HBI Senior Biomedical Engineer. J.M.B. provides knowledge of skull base anatomy advised on designing model as part of fellowship position, Princess Alexandra Hospital, Brisbane, at Queensland Health. M.W. is Clinical Director of HBI. HBI provides support with printing and drilling materials and equipment as well as engineering and design expertise. HBI received equipment on loan from Zeiss and from Medtronic. There is potential for HBI to produce 3D Printed models commercially in the future. Project Intellectual property is owned by Metro North Hospital and Health Service and The University of Queensland as tenants in common in shares proportionate to their respective inventive contribution to the development or creation of that Project IP. All authors received the equipment on loan from Zeiss, Medtronic and Herston Biofabrication. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Gold Coast University Hospital Human Research Ethics Committee (HREC/2022/QGC/87159). Informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. ASOHNS. Training and Education. Available online: https://asohns.org.au/Training-and-Education/SETProgram/Intermediate-Requirements
2. Chien WW, da Cruz MJ, Francis HW. Validation of a 3D-printed human temporal bone model for otology surgical skill training. World J Otorhinolaryngol Head Neck Surg 2021;7:88-93. [Crossref] [PubMed]
3. McMillan A, Kocharyan A, Dekker SE, et al. Comparison of Materials Used for 3D-Printing Temporal Bone Models to Simulate Surgical Dissection. Ann Otol Rhinol Laryngol 2020;129:1168-73. [Crossref] [PubMed]
4. Narang P, Raju B, Jumah F, et al. The Evolution of 3D Anatomical Models: A Brief Historical Overview. World Neurosurg 2021;155:135-43. [Crossref] [PubMed]
5. Frithioff A, Weiss K, Senn P, et al. 3D-printed temporal bone models for training: Does material transparency matter? Int J Pediatr Otorhinolaryngol 2024;184:112059. [Crossref] [PubMed]
6. Bone TM, Mowry SE. Content Validity of Temporal Bone Models Printed Via Inexpensive Methods and Materials. Otol Neurotol 2016;37:1183-8. [Crossref] [PubMed]
7. Cohen J, Reyes SA. Creation of a 3D printed temporal bone model from clinical CT data. Am J Otolaryngol 2015;36:619-24. [Crossref] [PubMed]
8. Crafts TD, Ellsperman SE, Wannemuehler TJ, et al. Three-Dimensional Printing and Its Applications in Otorhinolaryngology-Head and Neck Surgery. Otolaryngol Head Neck Surg 2017;156:999-1010. [Crossref] [PubMed]
9. Da Cruz MJ, Francis HW. Face and content validation of a novel three-dimensional printed temporal bone for surgical skills development. J Laryngol Otol 2015;129:S23-9. [Crossref] [PubMed]
10. Freiser ME, Ghodadra A, Hirsch BE, et al. Evaluation of 3D Printed Temporal Bone Models in Preparation for Middle Cranial Fossa Surgery. Otol Neurotol 2019;40:246-53. [Crossref] [PubMed]
11. Freiser ME, Ghodadra A, McCall AA, et al. Operable, Low-Cost, High-Resolution, Patient-Specific 3D Printed Temporal Bones for Surgical Simulation and Evaluation. Ann Otol Rhinol Laryngol 2021;130:1044-51. [Crossref] [PubMed]
12. Frithioff A, Frendø M, Weiss K, et al. Effect of 3D-Printed Models on Cadaveric Dissection in Temporal Bone Training. OTO Open 2021;5:2473974X211065012.
13. Frithioff A, Weiss K, Frendø M, et al. 3D-printing a cost-effective model for mastoidectomy training. 3D Print Med 2023;9:12.
14. George AP, De R. Review of temporal bone dissection teaching: how it was, is and will be. J Laryngol Otol 2010;124:119-25. [Crossref] [PubMed]
15. Gadaleta DJ, Huang D, Rankin N, et al. 3D printed temporal bone as a tool for otologic surgery simulation. Am J Otolaryngol 2020;41:102273. [Crossref] [PubMed]
16. Hochman JB, Rhodes C, Wong D, et al. Comparison of cadaveric and isomorphic three-dimensional printed models in temporal bone education. Laryngoscope 2015;125:2353-7. [Crossref] [PubMed]
17. Hong CJ, Giannopoulos AA, Hong BY, et al. Clinical applications of three-dimensional printing in otolaryngology-head and neck surgery: A systematic review. Laryngoscope 2019;129:2045-52. [Crossref] [PubMed]
18. Jenks CM, Patel V, Bennett B, et al. Development of a 3-Dimensional Middle Ear Model to Teach Anatomy and Endoscopic Ear Surgical Skills. OTO Open 2021;5:2473974X211046598.
19. McMillan A, Kocharyan A, Dekker SE, et al. Comparison of Materials Used for 3D-Printing Temporal Bone Models to Simulate Surgical Dissection. Ann Otol Rhinol Laryngol 2020;129:1168-73. [Crossref] [PubMed]
20. Kanaganayagam E, Bowman J, Forrestal D, et al. Air quality assessment of dissecting a J750 three-dimensional printed temporal bone and recommendations 2024. Aust J Otolaryngol 2025;8:5.
21. Mowry SE, Jabbour N, Rose AS, et al. Multi-institutional Comparison of Temporal Bone Models: A Collaboration of the AAO-HNSF 3D-Printed Temporal Bone Working Group. Otolaryngol Head Neck Surg 2021;164:1077-84. [Crossref] [PubMed]
22. Frithioff A, Frendø M, Pedersen DB, et al. 3D-Printed Models for Temporal Bone Surgical Training: A Systematic Review. Otolaryngol Head Neck Surg 2021;165:617-25. [Crossref] [PubMed]