RESUMO
Objective Current transnasal endoscopic techniques for sinus and skull base surgery use a single endoscope to provide visualization from one perspective curtailing depth perception and compromising visualization of the instrument-target interface. The view can be blocked by instruments, and collisions between instruments often occur. The objective of this study was to investigate the use of multiportal retrograde endoscopy to provide more accurate manipulation of the surgical target. Design Maxillary antrostomy and frontal sinusotomy were performed on three different cadavers by three different surgeons. A zero-degree rigid endoscope was introduced through the nose for the standard transnasal approach. A flexible endoscope was introduced transorally, directed past the palate superiorly, and then flexed 180 degrees for the retrograde view. Videos of the standard transnasal view from the rigid endoscope and retrograde view from the flexible endoscope were recorded simultaneously. Results All surgeries were able to be performed with dual-screen viewing of the standard and retrograde view. The surgeons noted that they utilized the retrograde view to adjust the location of ends/tips of their instruments. Four surgeons reviewed the videos and individually agreed that the visualization achieved provided a perspective otherwise not attainable with rigid transnasal endoscopy alone. Conclusion High-quality visualization of surgical targets such as the frontal or maxillary ostia can be challenging with rigid endoscopes alone. Multiportal retrograde endoscopy provides proof of concept that additional views of a surgical target can be achieved. Additional work is needed to further develop indications, techniques, and generalizability to targets beyond those investigated here.
RESUMO
Accurate instrument segmentation in the endoscopic vision of minimally invasive surgery is challenging due to complex instruments and environments. Deep learning techniques have shown competitive performance in recent years. However, deep learning usually requires a large amount of labeled data to achieve accurate prediction, which poses a significant workload. To alleviate this workload, we propose an active learning-based framework to generate synthetic images for efficient neural network training. In each active learning iteration, a small number of informative unlabeled images are first queried by active learning and manually labeled. Next, synthetic images are generated based on these selected images. The instruments and backgrounds are cropped out and randomly combined with blending and fusion near the boundary. The proposed method leverages the advantage of both active learning and synthetic images. The effectiveness of the proposed method is validated on two sinus surgery datasets and one intraabdominal surgery dataset. The results indicate a considerable performance improvement, especially when the size of the annotated dataset is small. All the code is open-sourced at: https://github.com/HaonanPeng/active_syn_generator.
RESUMO
Craniomaxillofacial vascular anomalies encompass a diverse and complex set of pathologies that may have a profound impact on pediatric patients. They are subdivided into vascular tumors and vascular malformations depending on biological properties, clinical course, and distribution patterns. Given the complexity and potential for leading to significant functional morbidity and esthetic concerns, a multidisciplinary approach is generally necessary to optimize patient outcomes. This article reviews the etiology, clinical course, diagnosis, and current management practices related to vascular anomalies in the head and neck.
Assuntos
Cabeça , Pescoço , Malformações Vasculares , Humanos , Criança , Malformações Vasculares/terapia , Malformações Vasculares/diagnóstico , Pescoço/irrigação sanguínea , Cabeça/irrigação sanguínea , Neoplasias de Cabeça e PescoçoRESUMO
OBJECTIVE: The objective of this study was to develop and assess multidisciplinary advanced surgical planning (ASP) sessions using three dimensional (3D) printed models for cervicothoracic slide tracheoplasty (CST). We hypothesized that these sessions would improve surgeon confidence, streamline intraoperative planning, and highlight the utility of 3D modeling. METHODS: 3D-printed patient-specific trachea models were used in pre-operative ASP sessions consisting of a multidisciplinary case discussion and hands-on slide tracheoplasty simulation. Participants completed a survey rating realism, utility, impact on the final surgical plan, and pre- and post-session confidence. Statistical analysis was performed via Wilcoxon and Kruskal-Wallis tests. RESULTS: Forty-eight surveys were collected across nine sessions and 27 different physicians. On a 5-point Likert scale, models were rated as "very realistic", "very useful" (both median of 4, IQR 3-4 and 4-5, respectively). Overall confidence increased by 1.4 points (+/- 0.7, p < 0.0001), with the largest change seen in those with minimal prior slide tracheoplasty experience (p = 0.005). Participants felt that the sessions "strongly" impacted their surgical plan or anticipated performance (median 4, IQR 4-5), regardless of training level or experience. CONCLUSION: 3D-printed patient-specific models were successfully implemented in ASP sessions for CST. Models were deemed very realistic and very useful by surgeons across multiple specialties and training levels. Surgical planning sessions also strongly impacted the final surgical plan and increased surgeon confidence for CST. LEVEL OF EVIDENCE: 4 Laryngoscope, 134:3395-3401, 2024.
Assuntos
Modelos Anatômicos , Impressão Tridimensional , Traqueia , Humanos , Traqueia/cirurgia , Procedimentos de Cirurgia Plástica/métodosRESUMO
Endoscopy is the gold standard for characterizing pediatric airway disorders, however, it is limited for quantitative analysis due to lack of three-dimensional (3D) vision and poor stereotactic depth perception. We utilize structure from motion (SfM) photogrammetry, to reconstruct 3D surfaces of pathologic and healthy pediatric larynges from monocular two-dimensional (2D) endoscopy. Models of pediatric subglottic stenosis were 3D printed and airway endoscopies were simulated. 3D surfaces were successfully reconstructed from endoscopic videos of all models using an SfM analysis toolkit. Average subglottic surface error between SfM reconstructed surfaces and 3D printed models was 0.65 mm as measured by Modified Hausdorff Distance. Average volumetric similarity between SfM surfaces and printed models was 0.82 as measured by Jaccard Index. SfM can be used to accurately reconstruct 3D surface renderings of the larynx from 2D endoscopy video. This technique has immense potential for use in quantitative analysis of airway geometry and virtual surgical planning.