Добавил:
kiopkiopkiop18@yandex.ru t.me/Prokururor I Вовсе не секретарь, но почту проверяю Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:

Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_537_Библиотеки_им_академика_М_И_Перельмана

.pdf
Скачиваний:
0
Добавлен:
29.08.2026
Размер:
86 Мб
Скачать
all neck levels [130]. A skin flap is elevated in this technique with the placement of a self-retaining retractor, creating a workspace for the robotic neck dissection. We refer to [24] the article for the details.
Additional research literature and the associated areas of TORS applications and observations included, but were not limited to, the following:
Tong neoplasms [137] Glottic microsurgery [138] Supraglottic partial laryngectomy [139] TORS of the oropharyngeal cancer [136, 140, 141, 143147]
4 Technology Integration into Robotics Use in Facial Plastic Surgery
Technologies integrated into robotics use in facial plastic surgery come along a long way within the last 10years. As is the case for dentistry and medicine, technology, when used prudently, can add the accuracy and speed in which oral and maxillofacial surgeons care for patients [15]. Technologies incorporate into robotics for plastic surgery procedures included (but are not limited to) CT scanning technologies, machine learning, deep learning, AR/VR, 3D simulation, computerized analysis, and natural language processing [105]. Even though these technologies can be beneficial to estimate an outcome from facial plastic surgery as well as offer assistance to the surgeon, the results shadow the actual outcomes noting that the difference between the predicted and actual outcomes can result in patient dissatisfaction [106, 107]. Machine learning in combination with perioperative photos can be advantageous over the robotic predictions [16].
Augmented reality has been incorporated into robotic-assisted craniomaxillofacial surgeries allowing ease of 3D simulation of the outcomes in [8] for the first time. A rapid prototype is introduced by the authors to position landmarks and robot-assisted arms to utilize an operative method. For comparative purposes, both traditional and robot-assisted procedures are tested. Toolkits of AR are used for system tracking and display with the operative plans and measured error carried out. Positioning and angle accuracy and stability results attained for the RAS method when compared to the traditional method indicated success of the combination of
https://t.me/medicina_free
AR and RAS use; Therefore, AR navigation use for mandibular angle split osteotomy with specialized robot-assisted arms was determined to be a success [8]. Figure 2 displays an example of an optical tracking system integrated into a Kuka robotic surgical system with a tool developed and used with the robot for dental applications [119].
Fig. 2 An integrated technological system that incorporates a robot and an optical tracker [119]
Another AR application to increase the effectiveness of mandible plastic surgery and reduce the burden of surgeons is developed in [9] for attaining comparative results of robotic-assisted and conventional plastic surgeries. As a part of this application, motor control followed fuzzy logic and AR assists surgeons with positioning. The concern associated with AR’s detection and position estimation of the outcomes of drilling is handled with the use of a force sensor. Experiments are conducted on animals for identifying the effectiveness of the robotic system. Errors associated with positioning and angles as well as sensor feedback and automatic drilling calculations for the designed system determined to result in accuracy of
https://t.me/medicina_free
positioning and automatic drilling that can result in improving the traditional plastic surgical procedures.
The robotics applications require training of surgeons and their skill assessments [31, 148].
Patient safety and operative performance can be improved by the assessment of objective metrics that relate to dissection and anastomosis [149]. Analysis of the data collected on visual and kinematic data captured during robotic surgeries provides opportunities to build effective tools; however, it comes with the associated challenges as the nature of the surgical motion data is complicated, hand motion generates variability in motion detection, and the data has a nonlinear nature [150]. Analysis of such data sets requires advanced analysis methods, such as machine learning.
Orbital bone positioning and visualization of orbital fractures is accomplished by using mixed reality in [16]. Noting that accurate localization of the orbital bone is critical for achieving orbital fracture reduction surgery (OFRS), a specific marker is designed and adopted by the authors to perform the registration step and the outcomes to be displayed in a platform of the navigation system. The precision is determined to be an advantage of the developed system while the need for localization accuracy improvement is mentioned by marker detection and registration methods, along with evaluation of subjective feedback on acceptance and usability [16].
Deep learning (DL), a subset of machine learning, serves to simplify the analysis of data sets with complicated nature that can be attained during facial plastic surgeries [13]. An application of DL that works in localized settings is region-based convolutional neural networks that help identify regions of interest with selective search to help localize objects of interest that resolve the recognition challenges [151]. One successful application of this method is the surgical tool and hand detection from surgical videos for extraction of more data [152]. Another useful method is attained by combining deep learning with reinforcement learning (called deep reinforcement learning) through enhancement of the machine’s knowledge by using a feedback mechanism to improve the reactions of the machine by using well-defined reward and punishment with the associated consequences applied. A combination of computer vision and deep
https://t.me/medicina_free
reinforcement learning is practically one application to enhance robotic­assisted surgeries [153].
In this application, entities within the intraoperative environment, such as organs and their movements, are read by a computer vision system while automatic motion recognition and pattern detection are analyzed by the deep reinforcement method [154].
The augmented and virtual reality applications require several steps to be taken during a reconstructive surgery as the surgery requires these steps. Initially, CT scan angiography data is used for loading segmented bone and vessels with resection of the bone to prepare the recipient site for reconstruction. Fibula segment positions, orientations, and angulations need to be defined and pedicle reach to anastomosis sites on the recipient vessels needs to be tested. Possible skin paddle configurations need to be figured out as well. Several iterations of the Design and Test stage would help to find a suitable configuration for the fibula, vessels, and skin paddle [108]. Figure 3 [109] displays the use of VR simulation using a haptic device.
https://t.me/medicina_free
Fig. 3 Simulation by using VR [110]
The ability to provide additional information without diminishing the camera’s information or giving a wrong perception is one of the challenges faced in AR applications. This limitation has been partially eliminated with the advancement of machine learning by using object subtraction method that had strong performance for the detection of instruments in laparoscopy videos, allowing the surgeon to view only the relevant information [112]. Figure 4 displays depth perception improvement methods for AR based on curvature-dependent transparency [111].
Fig. 4 Points chosen on a person for AR application with the view of how AR would demonstrate the skull area [111]
As far as the planning of VR application is concerned, the marking of the location is accomplished first, as shown on the left of Fig. 5. What follows next is the virtual cut of the marked location. The use of robotics for these two steps would be an advanced method of robotics and VR integration.
https://t.me/medicina_free
Fig. 5 The VR application developed in [109] for skull marking is displayed on the left image, and the image on the right side displays the next step with the removal of the marked region
The next step of the application requires raising skin flap and lifting anteriorly, as shown in the left image in Fig. 6. The haptic device stylus serves as the aneurysm clip holder in this image. The image on the right shows the clip holder and the clip that gives the 3D depth [109].
Fig. 6 The steps following Fig. 4 with a clip and clip holder displayed for the associated operation
Integration of robotics into VR in the abovementioned setting can be one of the new initiatives that have not been explored in the literature.
https://t.me/medicina_free
Augmented Reality is used for testing marker-free robotic surgical placement of two iliac crest transplants in [26]. The purpose of the AR and robotics integration was to determine the benefit of combining the two technologies for easing complex vascularized graft reconstructions and identifying their effectiveness in reducing facial skeleton–related defects. Figure 7 demonstrates a robot with a 3D camera that has a mini projector and a 3D camera attached to a mechanical system that allows the robot to move freely to operate on the projected iliac crest implant [113]. The key elements of the AR application and determining its effectiveness in [26] included random CT scans of two commonly used iliac crest transplantation model configurations, cutting guidelines following the surgical protocols, and the measurements of duration, accuracies of the distances, angulation, and the volume between the planned and executed osteotomies. The authors determined the time and accurate execution of preoperatively planned geometries to be the beneficial aspects of using, AR while anatomical complexities and vertical osteotomy were the weaknesses faced. It is concluded that the use of AR has benefits in applications with the shortcomings outlined above.
https://t.me/medicina_free
Fig. 7 A robot with a projector and 3D camera attached to a mechanical unit for easing the robot’s move to operate on an iliac crest model
Postoperative skeletal change prediction for orthognathic surgical planning by using machine learning is investigated in [29] and the authors showed its potential for reducing the workload of surgeons. Noting this success, the authors also pointed out its potential beneficial use with the robotic surgical hardware [114, 115].
Other technologies that are used alongside of robotics included computer-aided design and computer-aided manufacturing, which allowed symmetrical reconstruction of unilateral defects by mirroring the unaffected side [10, 160162], and simulation is used for demonstrating the surgical process with the associated lengths and angles calculated for fibular segments with marking of curves/lines [157, 158]; it is also possible to incorporate robotics as a part of these technologies themselves.
5 Free Flap Reconstruction
Free fibula flap is the typical application used for mandibular defect treatment. The application of robotics is seen in both the preoperative planning and during surgeries. In this section, we will cover a review of the literature that contains such applications.
Free fibula flap’s preoperative planning protocols are costly to learn and have execution challenges; these protocols are not robotic-assisted surgery friendly for cutting fibular osteotomy plane. Redesign of such preoperative surgical planning system is essential, which allows robotics assistance. Such a system is proposed in [10] with the associated adaptation of artificial intelligence with the incorporation of information attained from the CT scan images, as shown in Fig. 8 [10]. This figure displays the data read from the deformed mandibular structure with the associated CT scan images used for AI application.
https://t.me/medicina_free
Fig. 8 The progression starting with CT scanned images to demonstration of mandibular deformity within simulation and AI application
A part of the preplanning process is excision of defected mandible along with mirroring and repositioning of the planned mandible. This requires determination of an appropriate and accurate planar structure for the mandible, as shown in Fig. 9.
https://t.me/medicina_free
Fig. 9 The planar structure formed for mandibular reconstruction and the associated three dimensions
The planned mandibular structure is then incorporated into the simulation with the associated planned cutting of the bone selected for mandibular reconstruction displayed in Fig. 10 [10]. Path planning is then designed for robotic execution as a part of the operational planning using the landmarks determined for bone cutting and placement.
https://t.me/medicina_free