VR ASSISTED TEMPORAL BONE SURGERY AND TRAINING

been criticized, since it does not include some crucial surgical landmarks (e.g., antrum, facial recess) and does not assess the entire surgical process (Sethia et al., 2017).

While OSATS and the Welling scale and their modifications are

considered as the current leaders in the field, they are still claimed to need refinement and in overall large multicenter studies are still needed in order to increase the validity and feasibility of different scales prior to their universal acceptance (Sethia et al., 2017).

In recent years, surgical simulation training has introduced the

possibility for repeated and distributed training with immediate feedback or guidance and summative score at the final stage of the procedure.

Objective automated assessment tools for mastoidectomy have been under vigorous development in various VR simulator platforms (Al- Shahrestani et al., 2019), as it has been claimed that they could lead to significant time and resource savings in comparison to the traditional, and often subjective, assessment of surgical competence. However, there is no general consensus for their general acceptance; at present, there is

insufficient evidence of their validity and reliability and therefore further studies are necessary (Al-Shahrestani et al., 2019; Andersen et al., 2019).

The interaction with objects in VR environment is done with hand-held controllers. In addition, various haptic training instruments e.g.,

endoscopes, drills are being constantly developed. The positional tracking to synchronize the movements of the user’s head, controllers and the VR environment in real time, is usually accomplished using optical cameras and infrared laser triangulation and tracking markers (Sutherland et al., 2018).

The computer capacity, VR technologies and applications are all rapidly evolving and improvements in the feeling of immersion, convenience of hardware setup and accuracy of tracking are being developed by many device manufacturers.

The terminology surrounding different 3D computer generated environments is frequently confused and confusing. The terms VR, augmented reality (AR), mixed reality (MR) and extended reality (XR) are occasionally misused. In principle, in VR, the user exists in a fully virtual environment, without any intervention of the real world. With AR, virtual objects and information are overlapped in the field of view of the

surrounding real world without occluding the user’s view of the real world.

MR refers to a hybrid technology of VR and AR as it produces an

environment where the physical and digital objects and elements interact and coexist in real time. The term XR is usually considered as an umbrella term for all the aforementioned terms.

2.4.2History of VR

The history of VR is rather long; the first virtual concepts were developed already in the 1950s. In 1957, Morton Heilig introduced a machine called the Sensorama that was one of the earliest known examples of immersive multimodal technology including stereoscopic 3D display, stereo speakers, smells, wind and vibrations. It is considered to be one of the earliest VR systems created (“Heilig M. (1962). Sensorama simulator. U.S. Patent No - 3, 870. Virginia: United States Patent and Trade Office.,” n.d.). The first refined design of HMD is credited to Ivan Sutherland, who released a design called

“Sword of Damocles” in 1966.

The development of VR systems continued through the 1970s and as the rapid advances in computer graphics and technology continued in the 1980s, the first commercial VR applications with HMDs and data-gloves were introduced. The major force in VR technology development in the 1980s and 1990s was interest in flight training and simulation initiated by the National Aeronautics and Space Administration (NASA). Additionally, the hand-held controllers, data-gloves and other instruments enabling the manipulation of objects and giving haptic feedback were developed and refined. However, the high costs and relatively poor performance were still major problems. It took over two decades before a commercial

breakthrough occurred in modern HMDs and VR technology. The introduction of Oculus by Palmer Luckey in 2012 (Oculus, Menlo Park, California, United States) has led the way in the current era of VR

development and this breathed new life into this promising technology; it is estimated that its market value will reach 9.5 billion dollars by the year 2028 (“Augmented Reality & Virtual Reality In Healthcare Market Worth

$9.5 Billion By 2028 https://www.grandviewresearch.com/press- release/global-augmented-reality-ar-virtual-reality-vr-in-healthcare- market,” 2021).

2.4.3VR in medical use

The potential of VR technology for medical applications is attracting more and more interest; in fact the first medical VR applications were introduced in the 1990s. These mainly focused on the visualization of complex

anatomy, preoperative planning, surgery training simulations and telemedicine (Chinnock, 1994). Since then, the pros and cons of VR technology in medical education, surgery training, surgery planning, and visualization have been investigated (Cipresso et al., 2018; Li et al., 2021;

Zhang et al., 2021). In addition, the beneficial effects of VR in a wide range of therapeutic interventions e.g., physical and psychological rehabilitation, pain management, treatment of dementia and psychiatric disorders have been evaluated (Li et al., 2017). In recent years, the interest in medical VR

publications which have grown in the 2010s and early 2020s (Pensieri and Pennacchini, 2014; Yeung et al., 2021).

The use of VR in clinical medicine can be divided into three major parts:

training, surgery planning and medical image interpretation. In addition, applications designed directly for patient education and therapy (e.g., pain and anxiety management) are being tested and are topics of substantial research interest.

2.4.4VR in temporal bone surgical training

The earliest documented simulators for TB surgery were developed in the late 1990s and early 2000s (Kuppersmith et al., 1997; Wiet et al., 2000). The aim of these simulators was to provide a virtual TB dissection experience with comparable realism to live cadaveric dissection and this still is the aim of more modern devices. In principle, the anatomical data of simulated TBs are generated from various 3D anatomical atlases; the most refined

versions use a TB model remodeled from high-resolution cryosection images of a human TB (Sørensen et al., 2002; Trier et al., 2008).

In the last decades, various TB surgical simulators have been developed and refined reflecting the need for additional training methods (Sorensen et al., 2009; Wiet et al., 2009) and there are now many publications which support the benefits and validity of utilizing VR in the instruction of surgical procedures involving the TB, especially for training novice otologic

surgeons and residents (Arora et al., 2012a; Varoquier et al., 2017; Yi C.

Zhao et al., 2011). Another advantage of simulator training is the opportunity for repeated practice and the possibility to train at a

convenient time for the resident or otologic surgeon (Frendø et al., 2020).

Studies from other surgical fields (e.g., neurosurgery, interventional

radiology, laparoscopic surgery) support this conclusion (Cates et al., 2016;

Fiani et al., 2020; Mori et al., 2022).

In recent years, there has emerged the concept of self-directed learning without outside tutoring or supervision and this has now extended to training of TB surgical techniques. This approach is thought to benefit otologic surgeons in the beginning of their training and to spare valuable

cadaver and supervisor resources. A study with 20 medical students with no prior experience in TB dissection demonstrated that the self-directed and simulator-tutored VR training group achieved significantly higher dissection performance scores than the group of students receiving traditional training (Yi Chen Zhao et al., 2011).

However, there are studies that have not described similar results. A multicenter study in residents with varying levels of experience could not demonstrate a difference in dissection performance between VR and cadaveric TB training (Wiet et al., 2012). However, in a prospective study consisting of two cohorts of 20 otorhinolaryngology residents participating in a TB dissection course, it was demonstrated that the use of simulator- integrated-tutor function and on-screen step-by-step guide and illustration for a complete mastoidectomy in VR could lead to a 52% increase in

mastoidectomy performance in comparison to the traditional training method (Andersen et al., 2016). Additionally, it was noted that only two hours of self-directed VR simulation training was effective and that

mastoidectomy skills seemed to be transferable from the VR simulation to the traditional dissection setting (Andersen et al., 2016). A recent

prospective study with a controlled cohort, with 38 otorhinolaryngology residents demonstrated that even decentralized and self-directed

mastoidectomy training with a VR simulator prior to cadaveric dissection did improve residents´ performance scores by as much as 76% (Frendø et al., 2020).

2.4.5Patient-specific surgical planning and training in VR

Traditionally, training simulators are designed for educational purposes (e.g., basic surgical techniques). These tend to display simulated TBs with a standard anatomy but without any pathological processes. In the past decade, the possibility to rehearse and plan a patient-specific procedure in a virtual environment before undertaking any actual real-life surgery has attracted the interest of researchers and is under vigorous development (Tolsdorff et al., 2009).

Since the anatomical and pathological variations that are encountered in clinical work are missing in the simpler VR simulations, it is only when the possibility of patient-specific VR training and presurgical evaluation became an option that these approaches became attractive to many surgeons, especially those with more experience. A patient-specific VR simulation allows the surgeon to repeatedly try out different approaches and to make greater use of the patient´s CT or MRI images (Figure 10) in a similar way as would be done during actual surgery (Asit Arora et al., 2014;

Kashikar et al., 2019). It has been reported that an improved 3D understanding of the patient’s individual anatomy, the extent of the

pathology and the knowledge of the relationships of anatomical structures in 3D gained via the VR simulation could ultimately lead to improved surgical outcomes (Chan et al., 2016).

The key issues in many studies with patient-specific training are related to the laborious manual processing work of the radiological image data to delineate key anatomical landmarks but there are also concerns

surrounding adequate image quality without exposing patients to unsafe amounts of radiation. In other surgical fields e.g., in reconstructive surgery, there are studies demonstrating that virtual surgical planning may improve surgical accuracy and clinical outcomes (Mazzola et al., 2020; Shenaq and Matros, 2018).

The majority of the evidence supports the belief that VR simulations such as those used in training and the assessment of anatomy are beneficial. However, the clinical evidence to support their benefits in TB surgery is still scarce and only a few investigators have addressed patient- specific surgery planning and training with VR based simulators (Sethia and Wiet, 2015). In a prospective study, 16 otorhinolaryngology residents were first trained in the performance of mastoidectomy with a VR simulator with an individual cadaveric TB after which they then performed actual

dissection on the same bone. The study demonstrated a significant

increase in the surgeon’s confidence after the anatomic-specific VR training session with a positive correlation to their dissection performance (Locketz et al., 2017a). A recent study used the CBCT scans of patients undergoing

the exploitation of actual preoperative CBCT data in a patient-specific simulation was feasible and provided valuable information prior to the actual otologic surgery (S. A. W. Andersen et al., 2021).

Figure 10. Temporal bone in virtual reality (With courtesy of Adesante Oy, Turku, Finland)

2.4.6VR in radiological image interpretation

Recent advances in VR technology have raised interest in its use as a diagnostical tool. VR technology provides an opportunity to combine VR visualizations and radiological image data to achieve a more immersive 3D visualization of radiological images leading to improvements in an

understanding of the underlying pathologies and ultimately to a more accurate diagnosis (Muff et al., 2022). The patient-specific training already touches on this issue, however there are only a few studies which have examined whether VR is beneficial in radiological image interpretation for diagnostic purposes (Elsayed et al., 2020; King et al., 2016). To the best of

the author´s knowledge, there are no previous studies concentrating only on the diagnostics of pathology in the TB.

2.4.7Temporal bone anatomy education in VR

Traditionally, anatomy education has relied on lectures, 2D image representations (e.g., radiological images, anatomy textbooks and illustrations), cadaver dissections and observing or assisting during live surgical operations. The inadequacy of resources for traditional 3D training (e.g., cadaver dissections) may hinder effective learning of complex 3D anatomy (Bergman et al., 2011). Thus, medical students, residents and even otologic surgeons at the beginning of their career frequently experience difficulties in obtaining an in-depth understanding of 3D anatomy from viewing 2D illustrations (Triepels et al., 2020).

Previous studies with medical students have demonstrated the feasibility and value of exploiting 3D techniques and VR environments in teaching anatomy (Chen et al., 2020; Yi-Chun Du et al., 2020). There are rather few studies focusing solely on the anatomy of the ear and the TB. A study with 73 neurosurgery residents comparing a 3D virtual TB model and a standard 2D anatomy resource demonstrated the utility of the virtual model in teaching TB’s anatomy. The virtual model allowed these

neurosurgery residents to understand the anatomy more realistically and 90% of them preferred the virtual 3D model over its 2D counterpart (Morone et al., 2019). In another study, otorhinolaryngology residents and specialists regardless of their training level, assessed that the VR TB model would be useful in anatomic education (Yamazaki et al., 2021).

2.4.8Limitations and challenges of VR in temporal bone

Despite the rapid development of VR technology in recent years, some challenges are still encountered. The small anatomical structures (e.g., ossicles, round window membrane), complex neurovascular structures and their individual passage pose a challenge, especially when patient-specific or cadaveric image data is used (Asit Arora et al., 2014). In addition, the

and a realistic VR visualization is a time consuming and laborious process (Steven Arild Wuyts Andersen et al., 2021). In attempts to address these challenges, different automated segmentation softwares are being introduced and issues surrounding the insufficient image quality in VR surgery simulations as well as in VR planning are under vigorous

development (S. A. W. Andersen et al., 2021; Powell et al., 2019; Tolsdorff et al., 2009).

Another challenge is the haptics of surgical simulation. For example, the vibrotactile experience of drilling and use of suction provide information about bone and soft tissues are an important aspect of TB drilling and dissection. The VR TB dissection simulators are often accompanied by different kinds of haptic instruments which attempt to simulate the tactile feedback from drills and tissue (Hochman et al., 2014). Various TB surgery simulator studies have noted that it is challenging to achieve a precise and realistic simulation of force-feedback and this is frequently listed as a limitation by users and obviously requires development (Ghasemloonia et al., 2016; Kerwin et al., 2017; Varoquier et al., 2017). However, the benefits and the need for highly realistic haptics for high-quality surgical simulation have been disputed, especially in how much they will help residents and novice otologic surgeons at the beginning of their training and in situations of minimally invasive surgery (e.g., robot-assisted surgery) (Compton et al., 2020; van der Meijden and Schijven, 2009).

It is only few decades ago when the high costs of VR and computer technology were limiting the deployment of this technology in medical use but in the most recent decade, the costs of VR software, applications, HMDs and controllers have decreased dramatically. Studies concentrating on cost-effectiveness of VR training and surgery planning have

demonstrated a positive effect in terms of saving resources and time (Cohen et al., 2010; Mazzola et al., 2020). However, there are only a few of these kinds of studies involving VR technology in TB training, planning and education and the pros and cons of VR on costs and clinical outcomes in TB surgery need to be clarified (A. Arora et al., 2014; Deutsch et al., 2015).

2.4.9Adverse effects of VR

Some adverse effects of the use of VR applications have been described;

these have been called VR- or cybersickness (Stanney et al., 2003). The use of head mounted displays has been associated with nausea, oculomotor disturbances, disorientation and headache (Weech et al., 2019). The main reason for these symptoms is the movement of the virtual environment e.g., locomotion or the movements of objects that do not translate to real- life physical motion (Rebenitsch and Owen, 2016) which in some users leads to a vestibular mismatch and to an adverse reaction. In addition, the insufficient resolution and screen frame rate or refresh rate can lead to discomfort and nausea (Saredakis et al., 2020). The design of modern VR applications and screens is intended to minimize these adverse effects, but further studies will be needed to better understand this phenomenon.