Hasil untuk "Surgery"

Shuo Zhang, Zihua Wang, Changgeng He et al.

This paper proposes LiNUS, a lightweight deep learning framework for the automatic segmentation of the Subthalamic Nucleus (STN) in Deep Brain Stimulation (DBS) surgery. Addressing the challenges of small target volume and class imbalance in MRI data, LiNUS improves upon the U-Net architecture by introducing spectral normalization constraints, bilinear interpolation upsampling, and a multi-scale feature fusion mechanism. Experimental results on the Tsinghua DBS dataset (TT14) demonstrate that LiNUS achieves a Dice coefficient of 0.679 with an inference time of only 0.05 seconds per subject, significantly outperforming traditional manual and registration-based methods. Further validation on high-resolution data confirms the model's robustness, achieving a Dice score of 0.89. A dedicated Graphical User Interface (GUI) was also developed to facilitate real-time clinical application.

Zhenkun Li, Shunyu Wan, Hugo Zhou

For any knot $K$ in $S^3$ and any positive rational $r$, we show that smooth $(-r)$-surgery on $K$ always admits a tight contact structure. More specifically, the tightness is detected by the non-vanishing Heegaard Floer contact invariant.

Srikar Annamraju, Yuxi Chen, Jooyoung Lim et al.

Goal: A limitation in robotic surgery is the lack of force feedback, due to challenges in suitable sensing techniques. To enhance the perception of the surgeons and precise force rendering, estimation of these forces along with tissue deformation level is presented here. Methods: An experimental test bed is built for studying the interaction, and the forces are estimated from the raw data. Since tissue deformation and stiffness are non-linearly related, they are independently computed for enhanced reliability. A Convolutional Neural Network (CNN) based vision model is deployed, and both classification and regression models are developed. Results: The forces applied on the tissue are estimated, and the tissue is classified based on its deformation. The exact deformation of the tissue is also computed. Conclusions: The surgeons can render precise forces and detect tumors using the proposed method. The rarely discussed efficacy of computing the deformation level is also demonstrated.

Ha Na Cho, Sairam Sutari, Alexander Lopez et al.

Objective: To develop and evaluate machine learning (ML) models for predicting length of stay (LOS) in elective spine surgery, with a focus on the benefits of temporal modeling and model interpretability. Materials and Methods: We compared traditional ML models (e.g., linear regression, random forest, support vector machine (SVM), and XGBoost) with our developed model, SurgeryLSTM, a masked bidirectional long short-term memory (BiLSTM) with an attention, using structured perioperative electronic health records (EHR) data. Performance was evaluated using the coefficient of determination (R2), and key predictors were identified using explainable AI. Results: SurgeryLSTM achieved the highest predictive accuracy (R2=0.86), outperforming XGBoost (R2 = 0.85) and baseline models. The attention mechanism improved interpretability by dynamically identifying influential temporal segments within preoperative clinical sequences, allowing clinicians to trace which events or features most contributed to each LOS prediction. Key predictors of LOS included bone disorder, chronic kidney disease, and lumbar fusion identified as the most impactful predictors of LOS. Discussion: Temporal modeling with attention mechanisms significantly improves LOS prediction by capturing the sequential nature of patient data. Unlike static models, SurgeryLSTM provides both higher accuracy and greater interpretability, which are critical for clinical adoption. These results highlight the potential of integrating attention-based temporal models into hospital planning workflows. Conclusion: SurgeryLSTM presents an effective and interpretable AI solution for LOS prediction in elective spine surgery. Our findings support the integration of temporal, explainable ML approaches into clinical decision support systems to enhance discharge readiness and individualized patient care.

Fan Jiang, Honglin Yu, Grace Chung et al.

The capabilities of Large Language Models (LLMs) in low-resource languages lag far behind those in English, making their universal accessibility a significant challenge. To alleviate this, we present $\textit{Franken-Adapter}$, a modular language adaptation approach for decoder-only LLMs with embedding surgery. Our method begins by creating customized vocabularies for target languages and performing language adaptation through embedding tuning on multilingual data. These pre-trained embeddings are subsequently integrated with LLMs that have been instruction-tuned on English alignment data to enable zero-shot cross-lingual transfer. Our experiments on $\texttt{Gemma2}$ models with up to 27B parameters demonstrate improvements of up to 20% across 96 languages, spanning both discriminative and generative tasks, with minimal regressions ($<$1%) in English. Further in-depth analysis reveals the critical role of customizing tokenizers in enhancing language adaptation, while boosting inference efficiency. Additionally, we show the versatility of our method by achieving a 14% improvement over a math-optimized LLM across 20 languages, offering a modular solution to transfer reasoning abilities across languages post hoc.

Boyi Ma, Yanguang Zhao, Jie Wang et al.

The DeepSeek models have shown exceptional performance in general scene understanding, question-answering (QA), and text generation tasks, owing to their efficient training paradigm and strong reasoning capabilities. In this study, we investigate the dialogue capabilities of the DeepSeek model in robotic surgery scenarios, focusing on tasks such as Single Phrase QA, Visual QA, and Detailed Description. The Single Phrase QA tasks further include sub-tasks such as surgical instrument recognition, action understanding, and spatial position analysis. We conduct extensive evaluations using publicly available datasets, including EndoVis18 and CholecT50, along with their corresponding dialogue data. Our empirical study shows that, compared to existing general-purpose multimodal large language models, DeepSeek-VL2 performs better on complex understanding tasks in surgical scenes. Additionally, although DeepSeek-V3 is purely a language model, we find that when image tokens are directly inputted, the model demonstrates better performance on single-sentence QA tasks. However, overall, the DeepSeek models still fall short of meeting the clinical requirements for understanding surgical scenes. Under general prompts, DeepSeek models lack the ability to effectively analyze global surgical concepts and fail to provide detailed insights into surgical scenarios. Based on our observations, we argue that the DeepSeek models are not ready for vision-language tasks in surgical contexts without fine-tuning on surgery-specific datasets.

Michelle Mattille, Alexandre Mesot, Miriam Weisskopf et al.

Flexible robots hold great promise for enhancing minimally invasive surgery (MIS) by providing superior dexterity, precise control, and safe tissue interaction. Yet, translating these advantages into endoscopic interventions within open cavities remains challenging. The lack of anatomical constraints and the inherent flexibility of such devices complicate their control, while the limited field of view of endoscopes restricts situational awareness. We present a robotic platform designed to overcome these challenges and demonstrate its potential in fetoscopic laser coagulation, a complex MIS procedure typically performed only by highly experienced surgeons. Our system combines a magnetically actuated flexible endoscope with teleoperated and semi-autonomous navigation capabilities for performing targeted laser ablations. To enhance surgical awareness, the platform reconstructs real-time mosaics of the endoscopic scene, providing an extended and continuous visual context. The ability of this system to address the key limitations of MIS in open spaces is validated in vivo in an ovine model.

Yiting Wang, Yunxin Fan, Fei Liu

Accurate modeling of tool-tissue interactions in robotic surgery requires precise tracking of deformable tissues and integration of surgical domain knowledge. Traditional methods rely on labor-intensive annotations or rigid assumptions, limiting flexibility. We propose a framework combining sparse keypoint tracking and probabilistic modeling that propagates expert-annotated landmarks across endoscopic frames, even with large tissue deformations. Clustered tissue keypoints enable dynamic local transformation construction via PCA, and tool poses, tracked similarly, are expressed relative to these frames. Embedding these into a Task-Parameterized Gaussian Mixture Model (TP-GMM) integrates data-driven observations with labeled clinical expertise, effectively predicting relative tool-tissue poses and enhancing visual understanding of robotic surgical motions directly from video data.

Maximilian Fischer, Florian M. Hauptmann, Robin Peretzke et al.

Although advances in brain surgery techniques have led to fewer postoperative complications requiring Intensive Care Unit (ICU) monitoring, the routine transfer of patients to the ICU remains the clinical standard, despite its high cost. Predictive Gradient Boosted Trees based on clinical data have attempted to optimize ICU admission by identifying key risk factors pre-operatively; however, these approaches overlook valuable imaging data that could enhance prediction accuracy. In this work, we show that multimodal approaches that combine clinical data with imaging data outperform the current clinical data only baseline from 0.29 [F1] to 0.30 [F1], when only pre-operative clinical data is used and from 0.37 [F1] to 0.41 [F1], for pre- and post-operative data. This study demonstrates that effective ICU admission prediction benefits from multimodal data fusion, especially in contexts of severe class imbalance.

Morteza Rasouligandomani, Alex del Arco, Tomaso Villa et al.

Background: Proximal Junctional Failure (PJF) is a post-operative complication in adult spine surgery, often requiring reoperation. Osteotomy is often used in revision surgeries, leading to 34.8% complications. Hence, suboptimal decisions might be extending hardware without osteotomy, which yields to severe Global Alignment and Proportion (GAP) scores. High GAPs increase PJF risk, but Hardware Density Reduction (HDR) might limit it. Methods: Two clinical cases were evaluated: 1) Initially operated with hardware extended to T10, GAP 10; 2) PJF at T11 and hardware extended to T3, GAP 11. Two patient-personalized spine FE models were constructed through Statistical Shape Modelling (SSM) and mesh morphing. Intervertebral Disk (IVD) fiber strain, screw pull-out force, and rod stress were evaluated for the cases 1) and 2), also for 91 virtual HDR scenarios with different GAP scores, using Finite Element (FE) simulations. Different rod and bone material properties were also assessed. Results: HDR could decrease IVD fiber strain (-70% at most) and increase screw pull-out forces (+142% at most) for cases with Ti rod and normal bone. Cr-Co rod and osteopenia, and osteoporotic bones had high PJF risk. Trade-off analyses could determine the best configurations avoiding PJF. Virtual cohort study showed that GAP 12 and 13 could not avoid PJF in any HDR scenarios either with Ti or Cr-Co rods. HDR in a UIV T10 virtual patient with GAP 11 could not de-risk in case of Cr-Co rods. UIV T3 with GAP 13 could not benefit any HDR strategy, independently of rod properties. In contrast, Ti rods might allow HDR to de-risk GAP 12 patients with UIV T3. Conclusions: HDR could avoid PJF in the patients with medium high GAP scores, depending on the screw reduction pattern, and bone and rod material properties. Remarkably, HDR technique might avoid excessive spine surgeries and minimize the surgery cost.

Lianhao Yin, Yutong Ban, Jennifer Eckhoff et al.

Understanding and anticipating intraoperative events and actions is critical for intraoperative assistance and decision-making during minimally invasive surgery. Automated prediction of events, actions, and the following consequences is addressed through various computational approaches with the objective of augmenting surgeons' perception and decision-making capabilities. We propose a predictive neural network that is capable of understanding and predicting critical interactive aspects of surgical workflow from intra-abdominal video, while flexibly leveraging surgical knowledge graphs. The approach incorporates a hypergraph-transformer (HGT) structure that encodes expert knowledge into the network design and predicts the hidden embedding of the graph. We verify our approach on established surgical datasets and applications, including the detection and prediction of action triplets, and the achievement of the Critical View of Safety (CVS). Moreover, we address specific, safety-related tasks, such as predicting the clipping of cystic duct or artery without prior achievement of the CVS. Our results demonstrate the superiority of our approach compared to unstructured alternatives.

Yuning Zhou, Henry Badgery, Matthew Read et al.

Self-supervised learning (SSL) has potential for effective representation learning in medical imaging, but the choice of data augmentation is critical and domain-specific. It remains uncertain if general augmentation policies suit surgical applications. In this work, we automate the search for suitable augmentation policies through a new method called Dimensionality Driven Augmentation Search (DDA). DDA leverages the local dimensionality of deep representations as a proxy target, and differentiably searches for suitable data augmentation policies in contrastive learning. We demonstrate the effectiveness and efficiency of DDA in navigating a large search space and successfully identifying an appropriate data augmentation policy for laparoscopic surgery. We systematically evaluate DDA across three laparoscopic image classification and segmentation tasks, where it significantly improves over existing baselines. Furthermore, DDA's optimised set of augmentations provides insight into domain-specific dependencies when applying contrastive learning in medical applications. For example, while hue is an effective augmentation for natural images, it is not advantageous for laparoscopic images.

Nguyen Tai Thu, Pham Dang Hai, Nguyen Thi Kieu Ly et al.

Abstract Background Cirrhosis is associated with chronic cardiovascular dysfunction termed cirrhotic cardiomyopathy (CCM), characterized by myocardial hypertrophy and diastolic dysfunction. Detecting early cardiac changes is crucial, especially in patients undergoing liver transplantation. Objective: This study aims to evaluate left ventricular systolic function in cirrhotic patients undergoing liver transplantation using speckle-tracking echocardiography. Methods A prospective observational study was conducted involving 54 cirrhotic patients who underwent liver transplantation, along with 28 age- and sex-matched healthy controls. Echocardiography, including conventional and two-dimensional speckle tracking echocardiography (2D-STE), was performed at baseline and one-month post-transplantation. Results The mean age in the cirrhotic group was 52.2 ± 12.7 years, with no significant difference compared to the control group. Viral hepatitis was the predominant etiology of cirrhosis (68.6%). Conventional echocardiography did not reveal significant differences between groups in LV ejection fraction [62% (56–69) vs. 59% (56–62); p = 0.830]. However, in cirrhotic patients, 2D-STE demonstrated significantly lower LV global longitudinal strain (LV-GLS) [17.5 (15.5–19.1) vs 19.0 (18.0–19.7), p = 0.006]. Post-transplantation, conventional echocardiography indices remained unchanged, while 2D-STE showed remarkable improvement in LV function, with increased LV-GLS compared to pre-transplantation value. Conclusions 2D-STE is a valuable tool for detecting and monitoring left ventricular systolic dysfunction in liver cirrhosis patients, particularly following transplantation. While conventional echocardiography may not detect subtle changes, 2D-STE reveals improvements in LV function post-transplantation, emphasizing its role in assessing cirrhotic cardiomyopathy.

Risza Subiantoro, Margarita M Maramis, Nining Febriana et al.

Introduction: Postoperative depression is a condition of depressive effects in patients without symptoms of depressive mood that occurs a few weeks after surgery and persists for at least 2 weeks. It generally possesses the same symptoms as major depressive disorder. Review: Their difference is that surgery is the trigger of depression in postoperative depression cases. Postoperative depression is associated with increased patients’ morbidity and mortality, increased the risk of disease complications, reduced postoperative healing process, prolonged the duration of treatment, and reduced patients’ quality of life. Therefore, mental health conditions should always be assessed on patients after undergoing surgery. Postoperative depression therapy needs to consider the benefits of antidepressants and adequate pain management. Antidepressant considerations also need to consider interactions with other drugs. Psychotherapy and cognitive behavioral therapy are also useful in postoperative depression management. Conclusion: This review is aimed to give insight about postoperative depression, its importance, and how to treat it.

A. L. Akopov

The development of systemic antitumor treatment, targeted therapy and immune checkpoint inhibitor therapy (ICI) has changed the treatment outcome and prognosis for locally advanced non-small cell lung cancer (NSCLC). Modern recommendations provide for the division of such patients into potentially resectable and unresectable, and such division does not always correspond to stage IIIa, IIIb or IIIc of the disease. The treatment of resectable tumors is recommended to start with neoadjuvant chemoimmunotherapy (CIT), followed by surgical intervention aimed at radical tumor removal only in patients who respond to treatment. For unresectable tumors, the best approach is considered to be simultaneous chemoradiation therapy (CRT) followed by adjuvant ICI therapy, and the role of surgical interventions is limited to salvage surgery – residual tumor or local relapse removal after non-surgical self-treatment.This approach to the treatment of patients with stage III NSCLC is unlikely to fully satisfy specialists and patients – there is no complete consensus on the definition of the term «resectability»; the possibility of converting an unresect able tumor into a resectable one as a result of neoadjuvant treatment is denied, which is contrary to clinical practice. The objective of the article was the critical analysis of existing recommendations on the role of the surgical approach as a stage of complex treatment of stage III NSCLC from the surgeon’s point of view.

Xin Zhang, Ge Wang, Xiaoru Li et al.

Background. Gastric cancer (GC) is the most common malignant tumor and ranks third in the world. LncRNA H19 (H19), one of the members of lncRNA, is overexpressed in various tumors. However, many undetermined molecular mechanisms by which H19 promotes GC progression still need to be further investigated. Methodology. A series of experiments was used to confirm the undetermined molecular mechanism including wound healing and transwell assays. Key Results. In this study, a significant upregulation of H19 expression was detected in GC cells and tissues. The poor overall survival was observed in GC patient with high H19 expression. Overexpression of H19 promoted the migration of GC cells, while knockdown of H19 significantly inhibited cell migration. Moreover, miR-148a-3p had a certain negative correlation with H19. Luciferase reporter assay confirmed that H19 could directly bind to miR-148a-3p. As expected, miR-148a mimics inhibited cell migration and invasion induced by H19 overexpression. The above findings proved that H19 functions as a miRNA sponge and verified that miR-148a-3p is the H19-associated miRNA in GC. We also confirmed that SOX-12 expression was upregulated in GC patient’s samples. SOX-12 expression was positively correlated with expression of H19 and was able to directly bind to miR-148a-3p. Importantly, in vitro wound healing assay showed that knockout of SOX-12 could reverse the promoting effect of H19 overexpression on cell migration. Conclusion. In conclusion, H19 has certain application value in the diagnosis and prognosis of GC. Specifically, H19 accelerates GCs to migration and metastasis by miR-138a-3p/SOX-12 axis.

Ying Li, Yabin Guo, Peipei Zhao et al.

Abstract Objectives This study aims to identify predictors of knee osteoarthritis (KOA) risk in middle-aged population, construct and validate a nomogram for KOA in this demographic. Methods From June to December 2020, we conducted a cross-sectional survey on 5,527 middle-aged individuals from Changsha and Zhangjiajie cities in Hunan Province, selected using a stratified multi-stage random sampling method. Data collection involved a structured questionnaire encompassing general demographic, physical condition, and lifestyle behaviors dimensions. The dataset was randomly split into a training set (n = 3868) and a validation set (n = 1659) at a 7:3 ratio via computerized randomization. We analyzed the prevalence of self-reported KOA and identified its influencing factors using logistic regression. A nomogram was constructed based on these "three-dimensional" factors. Subsequent validation was conducted, and the nomogram's performance was further evaluated through ROC curves, C-index, Hosmer–Lemeshow test, and calibration curves. Results The self-reported prevalence of KOA in the middle-aged population was 11.4% (632/5527). The risk factor with the greatest impact is: diagnosed with osteoporosis(95% CI 2.269–3.568, OR = 2.845), followed by age between 51 to 60 years (95% CI 2.176–3.151, OR = 2.619), diagnosed with hypertension(95% CI 1.633–2.499, OR = 2.02), diagnosed with diabetes (OR = 1.689), ethnic Han Chinese (OR = 1.673), exercise according to physical condition (OR = 1.643), pay attention to keeping the knee joint warm (OR = 1.535), eating habits are mainly light vegetables (OR = 1.374), male gender (OR = 1.343), drink occasionally in small amounts (OR = 1.286); a higher level of education (OR = 0.477) and frequently or always apply an external or plaster to relieve symptoms after knee discomfort (OR = 0.377; OR = 0.385) are protective factors. The C-index of the training set model was 0.8107 (95% CI: 0.8102–0.8111), with a statistically significant area under the ROC curve (AUC = 0.818), and the calibration curve showed a good fit. The C-index for the validation set was 0.8124 (95% CI: 0.8109–0.8140), with an AUC of 0.812. The Hosmer–Lemeshow test resulted in a P-value of 0.46 (P ≥ 0.05)indicating good calibration of the model. Conclusion The three dimensions nomogram generated in this study was a valid and easy-to-use tool for assessing the risk of KOA in middle-aged population, and helped healthcare professionals to screen the high-risk population.

Martin Bäcke

The Chekanov-Eliashberg dg-algebra is an algebraic invariant of Legendrian submanifolds of contact manifolds, whose definition recently has been extended to singular Legendrians. We describe a way of constructing simpler models of this dg-algebra for singular Legendrian knots in $\mathbb{R}^{3}$, and give the first examples of singular knots for which the full cohomology can be computed. This includes the $A_{n}$-Legendrians. An important question in the study of the Chekanov-Eliashberg dg-algebra is to understand how its quasi-isomorphism type changes when the Legendrian undergoes Weinstein ribbon isotopy. By explicit computation we show that quite dramatic changes are possible and that among other things, Weinstein ribbon isotopic Legendrians can have Koszul dual dg-algebras. Along the way, we compute the cohomology of the Chekanov-Eliashberg dg-algebra in the boundary of a Weinstein surface. This finishes the proof of the Bourgeois-Ekholm-Eliashberg surgery formula in dimension two, which was missing from the literature.

Charles Guinn, Samuel Stein, Esin Tureci et al.

Facilitating the ability to achieve logical qubit error rates below physical qubit error rates, error correction is anticipated to play an important role in scaling quantum computers. While many algorithms require millions of physical qubits to be executed with error correction, current superconducting qubit systems contain only hundreds of physical qubits. One of the most promising codes on the superconducting qubit platform is the surface code, requiring a realistically attainable error threshold and the ability to perform universal fault-tolerant quantum computing with local operations via lattice surgery and magic state injection. Surface code architectures easily generalize to single-chip planar layouts, however space and control hardware constraints point to limits on the number of qubits that can fit on one chip. Additionally, the planar routing on single-chip architectures leads to serialization of commuting gates and strain on classical decoding caused by large ancilla patches. A distributed multi-chip architecture utilizing the surface code can potentially solve these problems if one can optimize inter-chip gates, manage collisions in networking between chips, and minimize routing hardware costs. We propose QuIRC, a superconducting Quantum Interface Routing Card for Lattice Surgery between surface code modules inside of a single dilution refrigerator. QuIRC improves scaling by allowing connection of many modules, increases ancilla connectivity of surface code lattices, and offers improved transpilation of Pauli-based surface code circuits. QuIRC employs in-situ Entangled Pair (EP) generation protocols for communication. We explore potential topological layouts of QuIRC based on superconducting hardware fabrication constraints, and demonstrate reductions in ancilla patch size by up to 77.8%, and in layer transpilation size by 51.9% when compared to the single-chip case.

Zu‐Xuan Chen, Liang Liang, He‐Qing Huang et al.

Abstract Background and Aim The biological functions and clinical implications of lysophosphatidylcholine acyltransferase 1 (LPCAT1) remain unclarified in gastric cancer (GC). The aim of the current study was to explore the possible clinicopathological significance of LPCAT1 and its perspective mechanism in GC tissues. Materials and Methods The protein expression and mRNA levels of LPCAT1 were detected from in‐house immunohistochemistry and public high‐throughput RNA arrays and RNA sequencing. To have a comprehensive understanding of the clinical value of LPCAT1 in GC, all enrolled data were integrated to calculate the expression difference and standard mean difference (SMD). The biological mechanism of LPCAT1 in GC was confirmed by computational biology and in vitro experiments. Migration and invasion assays were also conducted to confirm the effect of LPCAT1 in GC. Results Both protein and mRNA expression levels of LPCAT1 in GC were remarkably higher than those in noncancerous controls. Comprehensively, the SMD of LPCAT1 mRNA was 1.11 (95% CI = 0.86–1.36) in GC, and the summarized AUC was 0.85 based on 15 datasets containing 1727 cases of GC and 940 cases of non‐GC controls. Moreover, LPCAT1 could accelerate the invasion and migration of GC by boosting the neutrophil degranulation pathway and disturbing the immune microenvironment. Conclusion An increased level of LPCAT1 may promote the progression of GC.