Search Results (57)

Background/Objectives: Colorectal anastomotic leak (CAL) is one of the most severe postoperative complications in colorectal surgery, impacting patient morbidity and mortality. Current risk assessment methods rely on clinical and intraoperative factors, but no real-time predictive tool exists. This study aimed to develop an artificial intelligence model based on intraoperative laparoscopic recording of the anastomosis for CAL prediction. Methods: A convolutional neural network (CNN) was trained with annotated frames from colorectal surgery videos across three international high-volume centers (Instituto Português de Oncologia de Lisboa, Hospital das Clínicas de Ribeirão Preto, and Royal Liverpool University Hospital). The dataset included a total of 5356 frames from 26 patients, 2007 with CAL and 3349 showing normal anastomosis. Four CNN architectures (EfficientNetB0, EfficientNetB7, ResNet50, and MobileNetV2) were tested. The models’ performance was evaluated using their sensitivity, specificity, accuracy, and area under the receiver operating characteristic (AUROC) curve. Heatmaps were generated to identify key image regions influencing predictions. Results: The best-performing model achieved an accuracy of 99.6%, AUROC of 99.6%, sensitivity of 99.2%, specificity of 100.0%, PPV of 100.0%, and NPV of 98.9%. The model reliably identified CAL-positive frames and provided visual explanations through heatmaps. Conclusions: To our knowledge, this is the first AI model developed to predict CAL using intraoperative video analysis. Its accuracy suggests the potential to redefine surgical decision-making by providing real-time risk assessment. Further refinement with a larger dataset and diverse surgical techniques could enable intraoperative interventions to prevent CAL before it occurs, marking a paradigm shift in colorectal surgery. Full article

Background/Objectives: The aim of this case series is to evaluate the outcomes and safety of video-assisted mastectomy, illustrating the harmonious collaboration of oncologic and plastic surgery. This novel minimally invasive technique allows immediate prosthetic reconstruction and represents a cost-effective alternative to robotic breast surgery. Methods: Video-assisted, single-port nipple-sparing mastectomies were performed in patients with small to medium-sized breasts, followed by immediate direct-to-implant reconstruction with either prepectoral or dual plane implant placement. The patients’ electronic medical records were analyzed, including demographic characteristics, operative times and histopathology reports. Results: A total of 18 patients underwent successful video-assisted mastectomy, without conversion to traditional open procedure. Fifteen of the operations were risk-reducing mastectomies. Twelve patients had complementary procedures performed concurrently on the previously operated contralateral breast (delayed reconstruction/expander-to-implant exchange). Moreover, three patients benefited from additional minimally invasive techniques during the same surgery (prophylactic laparoscopic hysterectomy). Immediate breast reconstruction with polyurethane or microtextured breast implants up to 450 cc was performed, with satisfactory aesthetic outcomes and no cancer recurrences at 6 to 12 months postoperative follow-up. Early complications included transient hypercapnia, areolar congestion and cellulitis. No skin necrosis or implant-related complications were reported. The most frequently encountered late issues were contour irregularities. Conclusions: Video-assisted mastectomy facilitates the safe removal of proven pathologic or healthy breast tissue with minimal damage to the breast’s skin envelope, facilitating single-stage breast reconstruction. Full article

Uterine fibroid is one of the most common benign uterine diseases, affecting up to 70–80% of females of reproductive age. Whilst abdominal myomectomy has traditionally been a major uterine-sparing surgical intervention for its management, this is not without considerable technical challenges and the potential for multiple complications and morbidity. Since the introduction of video-assisted endoscopic surgery by Dr. Camran Nezhat in the 1980s, the development of minimally invasive approaches to myomectomy has accelerated rapidly worldwide. Whilst this offers numerous benefits for patients, laparoscopic myomectomy still carries implications for necessary expertise in surgical skill, intraoperative hemorrhage control, concern for future reproductive potential and risk of occult uterine malignancy. In this review article, we present the latest data regarding such aspects and offer our opinions on widely raised questions and existing contentions regarding myomectomy. We believe that minimally invasive myomectomy is a safe, efficient and beneficial approach to management in the hands of surgeons empowered with advanced knowledge, experience, and refined surgical skills. Full article

Cotyledonoid dissecting leiomyoma (CDL), also known as Sternberg tumour, is a rare variant of leiomyoma that can be easily mistaken for a malignant neoplasm on clinical and radiological examination, posing a diagnostic challenge for clinicians. Background: Although the tumour can extend to neighbouring organs, it typically does not invade them and is considered benign. Therefore, it is essential to recognise and differentiate this leiomyoma variant from other malignancies to avoid misdiagnosis and overtreatment. Methods: This report depicts a unique case of CDL misdiagnosed as an ovarian tumour in a woman in her late 50s with post-menopausal bleeding and pelvic pressure. We initially planned and proceeded with a diagnostic laparoscopy and laparoscopic oophorectomy of the right ovarian mass, during which an intraoperative surprise of a retroperitoneal mass was explored and subsequently biopsied. Results: The final histopathology confirmed the presence of the rare fibroid variant CDL. The accompanying surgical video is among the first to feature a laparoscopic surgery of CDL and details the intraoperative findings and laparoscopic resection techniques utilised in this case. Conclusions: Given its rarity and non-specific clinical and radiological findings, diagnosing CDL pre-operatively can be challenging. This case prompts recognition and awareness of CDL and highlights the importance of careful consideration of uncommon differential diagnoses and thorough intraoperative exploration, with the goal of preventing the misdiagnosis and, consequently, overtreatment of unknown masses. Full article

The integration of Context-Aware Systems (CASs) in Future Operating Rooms (FORs) aims to enhance surgical workflows and outcomes through real-time data analysis. CASs require accurate classification of surgical tools, enabling the understanding of surgical actions. This study proposes a novel deep learning approach for surgical tool classification based on combining convolutional neural networks (CNNs), Feature Fusion Modules (FFMs), Squeeze-and-Excitation (SE) networks, and Bidirectional long-short term memory (BiLSTM) networks to capture both spatial and temporal features in laparoscopic surgical videos. We explored different modeling scenarios with respect to the location and number of SE blocks for multi-label surgical tool classification in the Cholec80 dataset. Furthermore, we analyzed a single-label surgical tool classification model using a simplified and computationally less expensive architecture compared to the multi-label problem setting. The single-label classification model showed an improved overall performance compared to the proposed multi-label classification model due to the increased complexity of identifying multiple tools simultaneously. Nonetheless, our results demonstrated that the proposed CNN-SE-FFM-BiLSTM multi-label model achieved competitive performance to state-of-the-art methods with excellent performance in detecting tools with complex usage patterns and in minority classes. Future work should focus on optimizing models for real-time applications, and broadening dataset evaluations to improve performance in diverse surgical environments. These improvements are crucial for the practical implementation of such models in CASs, ultimately aiming to enhance surgical workflows and patient outcomes in FORs. Full article

Background: Laparoscopic surgery for endometriosis presents unique challenges due to the complexity of and variability in lesion appearances within the abdominal cavity. This study investigates the application of deep learning models for object detection in laparoscopic videos, aiming to assist surgeons in accurately identifying and localizing endometriosis lesions and related anatomical structures. A custom dataset was curated, comprising of 199 video sequences and 205,725 frames. Of these, 17,560 frames were meticulously annotated by medical professionals. The dataset includes object detection annotations for 10 object classes relevant to endometriosis, alongside segmentation masks for some classes. Methods: To address the object detection task, we evaluated the performance of two deep learning models—FasterRCNN and YOLOv9—under both stratified and non-stratified training scenarios. Results: The experimental results demonstrated that stratified training significantly reduced the risk of data leakage and improved model generalization. The best-performing FasterRCNN object detection model achieved a high average test precision of 0.9811 ± 0.0084, recall of 0.7083 ± 0.0807, and mAP50 (mean average precision at 50% overlap) of 0.8185 ± 0.0562 across all presented classes. Despite these successes, the study also highlights the challenges posed by the weak annotations and class imbalances in the dataset, which impacted overall model performances. Conclusions: In conclusion, this study provides valuable insights into the application of deep learning for enhancing laparoscopic surgical precision in endometriosis treatment. The findings underscore the importance of robust dataset curation and advanced training strategies in developing reliable AI-assisted tools for surgical interventions. The latter could potentially improve the guidance of surgical interventions and prevent blind spots occurring in difficult to reach abdominal regions. Future work will focus on refining the dataset and exploring more sophisticated model architectures to further improve detection accuracy. Full article

Artificial intelligence (AI) has shown its potential to advance applications in various medical fields. One such area involves developing integrated AI-based systems to assist in laparoscopic surgery. Surgical tool detection and phase recognition are key components to develop such systems, and therefore, they have been extensively studied in recent years. Despite significant advancements in this field, previous image-based methods still face many challenges that limit their performance due to complex surgical scenes and limited annotated data. This study proposes a novel deep learning approach for classifying and localizing surgical tools in laparoscopic surgeries. The proposed approach uses a self-supervised learning algorithm for surgical tool classification followed by a weakly supervised algorithm for surgical tool localization, eliminating the need for explicit localization annotation. In particular, we leverage the Bidirectional Encoder Representation from Image Transformers (BEiT) model for tool classification and then utilize the heat maps generated from the multi-headed attention layers in the BEiT model for the localizing of these tools. Furthermore, the model incorporates class weights to address the class imbalance issue resulting from different usage frequencies of surgical tools in surgeries. Evaluated on the Cholec80 benchmark dataset, the proposed approach demonstrated high performance in surgical tool classification, surpassing previous works that utilize both spatial and temporal information. Additionally, the proposed weakly supervised learning approach achieved state-of-the-art results for the localization task. Full article

Background: Artificial Intelligence (AI)-driven training systems are becoming increasingly important in surgical education, particularly in the context of laparoscopic suturing. This systematic review aims to assess the impact of AI on skill acquisition, long-term retention, and clinical performance, with a specific focus on the types of machine learning (ML) techniques applied to laparoscopic suturing training and their associated advantages and limitations. Methods: A comprehensive search was conducted across multiple databases, including PubMed, IEEE Xplore, Cochrane Library, and ScienceDirect, for studies published between 2005 and 2024. Following the PRISMA guidelines, 1200 articles were initially screened, and 33 studies met the inclusion criteria. This review specifically focuses on ML techniques such as deep learning, motion capture, and video segmentation and their application in laparoscopic suturing training. The quality of the included studies was assessed, considering factors such as sample size, follow-up duration, and potential biases. Results: AI-based training systems have shown notable improvements in the laparoscopic suturing process, offering clear advantages over traditional methods. These systems enhance precision, efficiency, and long-term retention of key suturing skills. The use of personalized feedback and real-time performance tracking allows learners to gain proficiency more rapidly and ensures that skills are retained over time. These technologies are particularly beneficial for novice surgeons and provide valuable support in resource-limited settings, where access to expert instructors and advanced equipment may be scarce. Key machine learning techniques, including deep learning, motion capture, and video segmentation, have significantly improved specific suturing tasks, such as needle manipulation, insertion techniques, knot tying, and grip control, all of which are critical to mastering laparoscopic suturing. Conclusions: AI-driven training tools are reshaping laparoscopic suturing education by improving skill acquisition, providing real-time feedback, and enhancing long-term retention. Deep learning, motion capture, and video segmentation techniques have proven most effective in refining suturing tasks such as needle manipulation and knot tying. While AI offers significant advantages, limitations in accuracy, scalability, and integration remain. Further research, particularly large-scale, high-quality studies, is necessary to refine these tools and ensure their effective implementation in real-world clinical settings. Full article

Surgical data analysis is crucial for developing and integrating context-aware systems (CAS) in advanced operating rooms. Automatic detection of surgical tools is an essential component in CAS, as it enables the recognition of surgical activities and understanding the contextual status of the procedure. Acquiring surgical data is challenging due to ethical constraints and the complexity of establishing data recording infrastructures. For machine learning tasks, there is also the large burden of data labelling. Although a relatively large dataset, namely the Cholec80, is publicly available, it is limited to the binary label data corresponding to the surgical tool presence. In this work, 15,691 frames from five videos from the dataset have been labelled with bounding boxes for surgical tool localisation. These newly labelled data support future research in developing and evaluating object detection models, particularly in the laparoscopic image data analysis domain. Full article

Background and Objectives: The objective of this study was to assess the efficient use of advanced energy devices by examining the impact of their usage frequency on surgical outcomes of total laparoscopic hysterectomies. Materials and Methods: A retrospective study was conducted between 2020 and 2023 by a single surgeon. The patients’ medical records and surgical videos were reviewed. Cases were categorized into three groups based on the frequency of usage of advanced energy devices: Group 1 (≤10 uses), Group 2 (11–20 uses), and Group 3 (≥21 uses). The differences in blood loss, surgery time, and surgical outcomes among these groups were analyzed. This study was conducted as a single-center retrospective analysis. It included 126 patients who underwent total laparoscopic hysterectomy and provided informed consent for video recording. To evaluate the usage of advanced energy devices, anonymized surgical videos were reviewed, and outcomes were analyzed based on the frequency of usage of advanced energy devices. Results: The time required for surgery differed significantly among the three groups (p = 0.006). However, no significant differences were observed in the changes in hemoglobin levels or estimated blood loss (p = 0.255 and 0.053, respectively). Additionally, the application of hemostatic agents, the need for intraoperative or postoperative transfusions, and the use of intravenous hemostatic agents postoperatively showed no notable variation. Complication rates, including rates of hematoma, urinary tract injury, gastrointestinal injury, and infections necessitating reoperation, were also comparable. Conclusions: The findings suggest that the prudent and strategic use of advanced energy devices, rather than their frequent application, may improve surgical efficiency without increasing the risk of complications. Full article

Inguinal–iliac–obturator lymph node dissection is essential in the treatment of patients with cutaneous melanoma exhibiting the clinical or radiological involvement of pelvic lymph nodes. The open procedure is associated with elevated mortality rates. Numerous minimally invasive approaches have been suggested to mitigate the impact of this surgery on the patient’s quality of life. The preliminary findings of robotic-assisted dissection have been documented in the literature. They demonstrate a decrease in potential issues linked to robotic-assisted treatments as compared to open or video laparoscopic methods. No implications have been reported for long-term oncological outcomes. The present study compares the outcomes in 64 patients with robotic procedures, 187 with videoscopic procedures, and 83 with open pelvic lymph node dissection (PLND). However, the quality of evidence is too low to draw any valid conclusions. The available literature shows that a robotic procedure is feasible and has similar complication rates and oncological outcomes to other methods. The reason for the shorter operative time is not clear, but is associated with lower hospital costs. It is probable that, from a surgeon’s point of view, robotic techniques offer several advantages over videoendoscopic techniques, such as three-dimensional imaging, ergonomic control, and tools that mimic human hand movements. Randomized controlled trials are necessary to validate the benefits of robotic inguinal–iliac–obturator lymph node dissection (RIIOL) compared to videoscopic and open procedures, but the recruitment rate is very low because of the restricted indications for lymph node dissection against the background of the continuously evolving system of therapy. Full article

Background: Since laparoscopic surgery became the gold standard for colorectal procedures, specific skills are required to achieve good outcomes. The best way to acquire basic and advanced skills and reach the learning curve plateau is by using dedicated simulators: box-trainers, video-trainers and virtual reality simulators. Laparoscopic skills training outside the operating room is cost-beneficial, faster and safer, and does not harm the patient. When compared to box-trainers, virtual reality simulators and cadaver models have no additional benefits. Several laparoscopic trainers available on the market as well as homemade box and video-trainers, most of them using plastic boxes and standard webcams, were described in the literature. The majority of them involve training on a flat surface without any anatomical environment. In addition to their demonstrated benefits, box-trainers which add anatomic details can improve the training quality and skills development of surgeons. Methods: We created a 3D-printed anatomic pelvi-trainer which offers a real-size narrow pelvic space environment for training. The model was created starting with a CT-scan performed on a female pelvis from the Anatomy Museum (Cluj-Napoca University of Medicine and Pharmacy, Romania), using Invesalius 3 software (Centro de Tecnologia da informação Renato Archer CTI, InVesalius open-source software, Campinas, Brazil) for segmentation, Fusion 360 with Netfabb software (Autodesk software company, Fusion 360 with Netfabb, San Francisco, CA, USA) for 3D modeling and a FDM technology 3D printer (Stratasys 3D printing company, Fortus 380mc 3D printer, Minneapolis, MN, USA). In addition, a metal mold for casting silicone valves was made for camera and endoscopic instruments ports. The trainer was tested and compared using a laparoscopic camera, a standard full HD webcam and “V-Box” (INTECH—Innovative Training Technologies, Milano, Italia), a dedicated hard paper box. The pelvi-trainer was tested by 33 surgeons with different qualifications and expertise. Results: We made a complete box-trainer with a versatile 3D-printed pelvi-trainer inside, designed for a wide range of basic and advanced laparoscopic skills training in the narrow pelvic space. We assessed the feedback of 33 surgeons regarding their experience using the anatomic 3D-printed pelvi-trainer for laparoscopic surgery training in the narrow pelvic space. Each surgeon tested the pelvi-trainer in three different setups: using a laparoscopic camera, using a webcam connected to a laptop and a “V-BOX” hard paper box. In the experiments that were performed, each participant completed a questionnaire regarding his/her experience using the pelvi-trainer. The results were positive, validating the device as a valid tool for training. Conclusions: We validated the anatomic pelvi-trainer designed by our team as a valuable alternative for basic and advanced laparoscopic surgery training outside the operating room for pelvic organs procedures, proving that it supports a much faster learning curve for colorectal procedures without harming the patients. Full article

A 79 year old female individual presented to the hospital and complained of 1 month melena and anemia due to chronic gastrointestinal bleeding because of cavernous hemangiomatosis of the small bowel. After undergoing an initial video laparoscopic jejunal–ileal resection surgery 7 days after first hospitalization, given the persistence of anemia, she underwent laparotomic duodenojejunal resection surgery again 2 months later. Multiple cavernous hemangiomatosis is a rare vascular disease (7–10% of all benign small bowel tumors), and it often manifests with bleeding, which may be occult or massive; more rarely, it manifests with intestinal occlusion or perforation. Diagnoses often require the use of multiple radiological and endoscopic methods; video capsule endoscopy has significantly increased the diagnostic rate. The gold standard of treatment is surgical resection, whenever possible, balancing the need for radicality with the possible metabolic consequences of massive small intestine resections. Full article

The severity and invasiveness of clinical outcomes from organic responses to trauma are influenced by individual, surgical, and anesthetic factors. A stress response elicits neuroendocrine and immune reactions that may lead to multi-organ dysfunction. The degree of neuroinflammatory reflex activation from trauma can increase pro-inflammatory cytokine production, leading to endothelial dysfunction, glycocalyx damage, neutrophil activation, and multisystem tissue destruction. A shift in patient treatment towards a neuroinflammatory perspective has prompted a new evaluation protocol for surgical patients, required to understand surgical pathogenesis and its link to chosen anesthetic–surgical methods. The goal of this study is to summarize and disseminate the present knowledge about the mechanisms involved in immune and neuroendocrine responses, focusing on video laparoscopic surgeries. This article outlines various measures cited in the literature aimed at reducing the burden of surgical trauma. It reviews anesthetic drugs, anesthetic techniques, and intensive care procedures that are known to have immunomodulatory effects. The results show a preference for more sensitive inflammatory mediators to tissue trauma serving as care tools, indicators for prognosis, and therapeutic outcomes. Full article

Performing a minimally invasive surgery comes with a significant advantage regarding rehabilitating the patient after the operation. But it also causes difficulties, mainly for the surgeon or expert who performs the surgical intervention, since only visual information is available and they cannot use their tactile senses during keyhole surgeries. This is the case with laparoscopic hysterectomy since some organs are also difficult to distinguish based on visual information, making laparoscope-based hysterectomy challenging. In this paper, we propose a solution based on semantic segmentation, which can create pixel-accurate predictions of surgical images and differentiate the uterine arteries, ureters, and nerves. We trained three binary semantic segmentation models based on the U-Net architecture with the EfficientNet-b3 encoder; then, we developed two ensemble techniques that enhanced the segmentation performance. Our pixel-wise ensemble examines the segmentation map of the binary networks on the lowest level of pixels. The other algorithm developed is a region-based ensemble technique that takes this examination to a higher level and makes the ensemble based on every connected component detected by the binary segmentation networks. We also introduced and trained a classic multi-class semantic segmentation model as a reference and compared it to the ensemble-based approaches. We used 586 manually annotated images from 38 surgical videos for this research and published this dataset. Full article