A cadaver study of mastoidectomy using an image-guided human-robot collaborative control system

A Deep Learning Algorithm to Identify Anatomical Landmarks on Computed Tomography of the Temporal Bone

BACKGROUND

Petrous temporal bone cone-beam computed tomography scans help aid diagnosis and accurate identification of key operative landmarks in temporal bone and mastoid surgery. Our primary objective was to determine the accuracy of using a deep learning convolutional neural network algorithm to augment identification of structures on petrous temporal bone cone-beam computed tomography. Our secondary objective was to compare the accuracy of convolutional neural network structure identification when trained by a senior versus junior clinician.

METHODS

A total of 129 petrous temporal bone cone-beam computed tomography scans were obtained from an Australian public tertiary hospital. Key intraoperative landmarks were labeled in 68 scans using bounding boxes on axial and coronal slices at the level of the malleoincudal joint by an otolaryngology registrar and board-certified otolaryngologist. Automated structure identification was performed on axial and coronal slices of the remaining 61 scans using a convolutional neural network (Microsoft Custom Vision) trained using the labeled dataset. Convolutional neural network structure identification accuracy was manually verified by an otolaryngologist, and accuracy when trained by the registrar and otolaryngologist labeled datasets respectively was compared.

RESULTS

The convolutional neural network was able to perform automated structure identification in petrous temporal bone cone-beam computed tomography scans with a high degree of accuracy in both axial (0.958) and coronal (0.924) slices (P < .001). Convolutional neural network accuracy was proportionate to the seniority of the training clinician in structures with features more difficult to distinguish on single slices such as the cochlea, vestibule, and carotid canal.

CONCLUSION

Convolutional neural networks can perform automated structure identification in petrous temporal bone cone-beam computed tomography scans with a high degree of accuracy, with the performance being proportionate to the seniority of the training clinician. Training of the convolutional neural network by the most senior clinician is desirable to maximize the accuracy of the results.

Patient Benefit and Quality of Life after Robot-Assisted Head and Neck Surgery

Robotic systems for head and neck surgery are at different stages of technical development and clinical application. Currently, robotic systems are predominantly used for transoral surgery of the pharynx and larynx. Robotic surgery of the neck, the thyroid, and the middle and inner ear is much less common; however, some oncological and functional outcomes have been reported. This article provides an overview of the current state of robot-assisted head and neck surgery with a special emphasis on patient benefit and postoperative quality of life (QoL). The focus is placed on the role of transoral robotic surgery (TORS) for the resection of oropharyngeal carcinomas. For this application, reported long-term outcomes show functional post-operative advantages for selected oropharyngeal cancer patients after TORS compared to open surgery and primary radiotherapy. Since TORS also plays a significant role in the context of potential therapy de-escalation for HPV-positive oropharyngeal cancer patients, ongoing trials are presented. Regarding the evaluation of the therapeutic benefit and the QoL of cancer patients, special attention has to be paid to the large degree of variability of individual patients' preferences. Influencing factors and tools for a detailed assessment of QoL parameters are therefore detailed at the beginning of this article. Notably, while some robotic systems for ear and skull base surgery are being developed in Europe, TORS systems are mainly used in North America and Asia. In Europe and Germany in particular, transoral laser microsurgery (TLM) is a well-established technology for transoral tumor resection. Future trials comparing TORS and TLM with detailed investigation of QoL parameters are therefore warranted and might contribute to identifying suitable fields for the application of the different techniques.

Robotic Cochlear Implant Surgery: Imaging-Based Evaluation of Feasibility in Clinical Routine

Background: Robotic surgery has been proposed in various surgical fields to reduce recovery time, scarring, and to improve patients' outcomes. Such innovations are ever-growing and have now reached the field of cochlear implantation. To implement robotic ear surgery in routine, it is of interest if preoperative planning of a safe trajectory to the middle ear is possible with clinically available image data.

Methods: We evaluated the feasibility of robotic cochlear implant surgery in 50 patients (100 ears) scheduled for routine cochlear implant procedures based on clinically available imaging. The primary objective was to assess if available high-resolution computed tomography or cone beam tomography imaging is sufficient for planning a trajectory by an otological software. Secondary objectives were to assess the feasibility of cochlear implant surgery with a drill bit diameter of 1.8 mm, which is the currently used as a standard drill bit. Furthermore, it was evaluated if feasibility of robotic surgery could be increased when using smaller drill bit sizes. Cochlear and trajectory parameters of successfully planned ears were collected. Measurements were carried out by two observers and the interrater reliability was assessed using Cohen's Kappa.

Results: Under the prerequisite of the available image data being sufficient for the planning of the procedure, up to two thirds of ears were eligible for robotic cochlear implant surgery with the standard drill bit size of 1.8 mm. The main reason for inability to plan the keyhole access was insufficient image resolution causing anatomical landmarks not being accurately identified. Although currently not applicable in robotic cochlear implantation, narrower drill bit sizes ranging from 1.0 to 1.7 mm in diameter could increase feasibility up to 100%. The interrater agreement between the two observers was good for this data set.

Discussion: For robotic cochlear implant surgery, imaging with sufficient resolution is essential for preoperative assessment. A slice thickness of <0.3 mm is necessary for trajectory planning. This can be achieved by using digital volume tomography while radiation exposure can be kept to a minimum. Furthermore, surgeons who use the software tool, should be trained on a regular basis in order to achieve planning consistency.

Robotics for Cochlear Implantation Surgery: Challenges and Opportunities

OBJECTIVES

Recent advancements in robotics have set forth a growing body of evidence for the clinical application of the robotic cochlear implantation (RCI), with many potential benefits. This review aims to summarize these efforts, provide the latest developments in this exciting field, and explore the challenges associated with the clinical implementation of RCI.

DATA SOURCES

MEDLINE, PubMed, and EMBASE databases.

STUDY SELECTION

A search was conducted using the keywords "robotics otolaryngology," "robotic cochlear implant," "minimally-invasive cochlear implantation," "minimally-invasive mastoidectomy," and "percutaneous cochlear implant" with all of their synonyms. Literature selection criteria included articles published in English, and articles from 1970 to present.

RESULTS

The use of robotics in neurotology is a relatively new endeavor that continues to evolve. Robotics is being explored by various groups to facilitate in the various steps of cochlear implant surgery, including drilling a keyhole approach to the middle ear for implants, inner ear access, and electrode insertion into the cochlea. Initial clinical trials have successfully implanted selected subjects using robotics.

CONCLUSIONS

The use of robotics in cochlear implants remains in its very early stages. It is hoped that robotics will improve clinical outcomes. Although successful implants with robots are reported in the literature, there are some challenges that need to be addressed before this approach can become an acceptable option for the conventional cochlear implant surgery, such as safety, time, efficiency, and cost. However, it is hoped that further advancements in robotic technology will help in overcoming these barriers leading to successful implementation for clinical utility.

Robotics in otorhinolaryngology - head and neck surgery

A look at the past, present and future.