Local Segmentation Research Articles

Discriminative Segment Focus Network for Fine-grained Video Action Recognition

Fine-grained video action recognition aims at identifying minor and discriminative variations among fine categories of actions. While many recent action recognition methods have been proposed to better model spatio-temporal representations, how to model the interactions among discriminative atomic actions to effectively characterize inter-class and intra-class variations has been neglected, which is vital for understanding fine-grained actions. In this work, we devise a Discriminative Segment Focus Network (DSFNet) to mine the discriminability of segment correlations and localize discriminative action-relevant segments for fine-grained video action recognition. Firstly, we propose a hierarchic correlation reasoning (HCR) module which explicitly establishes correlations between different segments at multiple temporal scales and enhances each segment by exploiting the correlations with other segments. Secondly, a discriminative segment focus (DSF) module is devised to localize the most action-relevant segments from the enhanced representations of HCR by enforcing the consistency between the discriminability and the classification confidence of a given segment with a consistency constraint. Finally, these localized segment representations are combined with the global action representation of the whole video for boosting final recognition. Extensive experimental results on two fine-grained action recognition datasets, i.e., FineGym and Diving48, and two action recognition datasets, i.e., Kinetics400 and Something-Something, demonstrate the effectiveness of our approach compared with the state-of-the-art methods.

Re-resection Rates and Disease Recurrence in Crohn's Disease - A Population-Based Study Using Individual-Level Patient Data.

Despite advances in the medical treatment of Crohn's disease (CD), many patients will still need bowel resections and face the subsequent risk of recurrence and re-resection. We describe contemporary re-resection rates and identify disease-modifying factors and risk factors for re-resection. We conducted a retrospective, population-based, individual patient data cohort study covering 47.4% of the Danish population, including all CD patients who underwent a primary resection between 2010 and 2020. Among 631 primary resected patients, 24.5% underwent a second resection, and 5.3% a third. Re-resection rates after one, five, and 10 years were 12.6%, 22.4%, and 32.2%, respectively. Reasons for additional resections were mainly disease activity (57%) and stoma reversal (40%). Disease activity-driven re-resection rates after one, five, and 10 years were 3.6%, 10.1%, and 14.1%, respectively. Most stoma reversals occurred within one year (80%). The median time to recurrence was 11.0 months. Biologics started within one year of the first resection revealed protective effect against re-resection for stenotic and penetrating phenotypes. Prophylactic biologic therapy at primary ileocecal resection reduced disease recurrence and re-resection risk (HR 0.58, 95% CI (0.34-0.99), p=0.047). Risk factors for re-resection were location of resected bowel segments at the primary resection, disease location, disease behavior, smoking, and perianal disease. Re-resection rates, categorized by disease activity, are lower than those reported in other studies and are closely associated with disease phenotype and localization. Biological therapy may be disease-modifying for certain subgroups when initiated within one year of resection.

Actin polymerization and longitudinal actin fibers in axon initial segment plasticity.

The location of the axon initial segment (AIS) at the junction between the soma and axon of neurons makes it instrumental in maintaining neural polarity and as the site for action potential generation. The AIS is also capable of large-scale relocation in an activity-dependent manner. This represents a form of homeostatic plasticity in which neurons regulate their own excitability by changing the size and/or position of the AIS. While AIS plasticity is important for proper functionality of AIS-containing neurons, the cellular and molecular mechanisms of AIS plasticity are poorly understood. Here, we analyzed changes in the AIS actin cytoskeleton during AIS plasticity using 3D structured illumination microscopy (3D-SIM). We showed that the number of longitudinal actin fibers increased transiently 3 h after plasticity induction. We further showed that actin polymerization, especially formin mediated actin polymerization, is required for AIS plasticity and formation of longitudinal actin fibers. From the formin family of proteins, Daam1 localized to the ends of longitudinal actin fibers. These results indicate that active re-organization of the actin cytoskeleton is required for proper AIS plasticity.

Simplified Volumetric Models as an Effective Strategy for Segmenting Actin Networks in Cryo-Electron Tomograms.

Efficient methods for the extraction of features of interest remain one of the biggest challenges for the interpretation of cryo-electron tomograms. Various automated approaches have been proposed, many of which work well for high-contrast datasets where the features of interest can be easily detected and are clearly separated from one another. Our inner ear stereocilia cryo-electron tomographic datasets are characterized by a dense array of hexagonally packed actin filaments that are frequently cross-connected. These features make automated segmentation very challenging, further aggravated by the high-noise environment of cryo-electron tomograms and the high complexity of the densely packed features. Using prior knowledge about the actin bundle organization, we have placed layers of a highly simplified ball-and-stick actin model to first obtain a global fit to the density map, followed by regional and local adjustments of the model. We show that volumetric model building not only allows us to deal with the high complexity, but also provides precise measurements and statistics about the actin bundle. Volumetric models also serve as anchoring points for local segmentation, such as in the case of the actin-actin cross connectors. Volumetric model building, particularly when further augmented by computer-based automated fitting approaches, can be a powerful alternative when conventional automated segmentation approaches are not successful.

DEM-CFD Simulation Analysis of Heat Transfer Characteristics for Hydrogen Flow in Randomly Packed Beds

In this study, three randomly packed beds with varying tube-to-particle diameter ratios (D/d) are constructed using the discrete element method (DEM) and simulated via CFD under low pore Reynolds numbers (Rep < 100). An innovation of this research lies in the application of hydrogen in randomly packed beds, coupled with the consideration of its temperature-dependent thermal properties. The axial analysis of the heat transfer characteristics shows that PB−5 and PB−6 achieve thermal equilibrium 44% and 58% faster than PB−4, respectively, demonstrating enhanced heat transfer efficiency. However, at higher flow rates (0.8 m/s), the large-sized fluid channels in PB−6 severely impact the heat transfer efficiency, slightly reducing it compared to PB−5. Additionally, this study introduces a localized segmentation method for calculating the axial local Nusselt number, showing that the axial local Nusselt number (Nu) not only exhibits an inverse relationship with the axial porosity distribution, but also matches its amplitude fluctuations. The wall effect significantly impacts the flow and temperature distribution in the packed bed, causing notable velocity and temperature oscillations in the near-wall region. In the near-wall region, the average temperature is lower than in the core region, and the axial temperature profile exhibits more intense oscillations. These findings may provide insights into the use of hydrogen in randomly packed beds, which are vital for enhancing industrial applications such as hydrogen storage and utilization.

Gender differences in patterns of cartilage loss: Data from the Osteoarthritis Initiative

OBJECTIVEUnderstanding gender-specific differences in patterns of cartilage loss can improve our knowledge of the pathogenesis of knee osteoarthritis (KOA) development and progression and may inform clinical trials of treatments for KOA. The goal of our observational study was to examine gender differences in patterns of cartilage loss in the central weight-bearing regions of the femur. METHODSWe measured cartilage volume change in the indexed knee of 700 subjects with Kellgren-Lawrence (KL) 1, 2, or 3 from the OAI for four follow-up periods (BL to 24mo, BL to 48mo, BL to 72mo, and BL to 96mo) using the local area cartilage segmentation (LACS) method. Briefly, the LACS method uses robust coordinate systems fixed to anatomical landmarks to measure patterns of change in cartilage volume in sub-regions using responsiveness heat maps. RESULTSWe observed a statistically significant gender difference in cartilage change in the medial femur (MF), lateral femur (LF), and medial tibia (MT). The heat maps, showed loss was primarily in the posterior central weight-bearing portion of the LF and more general in the LT and MF. Similar patterns were observed for each of the four follow-up periods. CONCLUSIONSThe LACS method was capable of illustrating gender-specific differences in patterns of cartilage loss that may offer insight into the variation of gender differences in the natural history of KOA and may be useful in evaluating the benefit of interventions for KOA.

The influence of shear surface quality on the mechanical properties of long-span prestressed continuous rigid-frame bridge

PurposeThe long-span continuous rigid-frame bridges are commonly constructed by the section-by-section symmetrical balance suspension casting method. The deflection of these bridges is increasing over time. Wet joints are a typical construction feature of continuous rigid-frame bridges and will affect their integrity. To investigate the sensitivity of shear surface quality on the mechanical properties of long-span prestressed continuous rigid-frame bridges, a large serviced bridge is selected for analysis.Design/methodology/approachIts shear surface is examined and classified using the damage measuring method, and four levels are determined statistically based on the core sample integrity, cracking length and cracking depth. Based on the shear-friction theory of the shear surface, a 3D solid element-based finite element model of the selected bridge is established, taking into account factors such as damage location, damage number and damage of the shear surface. The simulated results on the stress distribution of the local segment, the shear surface opening and the beam deflection are extracted and analyzed.FindingsThe findings indicate that the main factors affecting the ultimate shear stress and shear strength of the shear surface are size, shear reinforcements, normal stress and friction performance of the shear surface. The connection strength of a single or a few shear surfaces decreases but with little effect on the local stress. Cracking and opening mainly occur at the 1/4 span. Compared with the rigid “Tie” connection, the mid-span deflection of the main span increases by 25.03% and the relative deflection of the section near the shear surface increases by 99.89%. However, when there are penetrating cracks and openings in the shear surface at the 1/2 span, compared with the 1/4 span position, the mid-span deflection of the main span and the relative deflection of the cross-section increase by 4.50%. The deflection of the main span increases with the failure of the shear surface.Originality/valueThese conclusions can guide the analysis of deflection development in long-span prestressed continuous rigid-frame bridges.

Novel deep learning methods for 3D flow field segmentation and classification

Flow field segmentation and classification help researchers to understand vortex structure and thus turbulent flow. Existing deep learning methods are mainly based on global vorticity information and focused on 2D circumstance. In this paper, based on flow field theory, criteria for flow field segmentation and classification in 3D space using local velocity information and the physical relationship between local vorticity and vortex wake are conducted, and then novel physical local information-embedded flow field segmentation and classification methods in 3D space using Multilayer Perceptron (MLP) are proposed. Our methods consist of three modules: a preprocessing module, a flow field segmentation module and a flow field classification module. The preprocessing module calculates input for succeeding modules based on the above criteria. The segmentation and classification modules process corresponding input to identify vortex structure, and further classify the type of vortex wakes in a 3D flow field. Experimental results show that our method can identify vortex area with an accuracy rate that reaches 90% while reducing time consumption by more than 90% compared with existing segmentation methods; our classification method reaches an accuracy rate of 99.7% while reducing time consumption by more than 90% compared with existing classification methods. Our proposed methods proved to achieve good performance both on accuracy and efficiency.

Digitalization era: Investigating the spatial interplay between cyber human activity and economy with a hierarchical framework

The development of Information and Communication Technology has shifted human activity from offline to online, and promoted digital economy. These changes challenge traditional methodologies relying on physical human activity as a micro-level reflection of the macro-level economy. To address this, a hierarchical framework is proposed to characterize cyber human activity, incorporating activity diversity, size, and preference. Then, Ordinary Least Squares and Geographically Weighted Regression models are used to examine the spatial interplay between cyber human activity and township-level economy. K-Medoids clustering is further applied to coefficients in cyber GWR model to reveal mechanisms influencing local economies. Taking diverse geospatial data in Jilin Province, China as an example, the result indicates that the significant extreme values of cyber human activity have divided Jilin Province into four subregions, closely aligning with local economic segmentation. Moreover, the local economy can be better reflected by cyber human activity rather than physical ones, especially in highly digitized regions. Furthermore, the regions with similar influence mechanisms cluster geographically. These clusters are categorized into three types, namely, dynamics, robust, and balanced, representing different mechanisms influencing the local economy. Practical and theoretical implications are discussed, including using cyber human activity in assessing the economy and implementing adaptable economic policies at township level.

Arterial Lesion Location and Outcomes of Intracranial Atherosclerotic Disease.

Intracranial atherosclerosis is a leading cause of stroke with a high recurrence rate despite treatment. Numerous factors are proposed to influence stroke recurrence due to intracranial atherosclerosis including lesion eccentricity, plaque characteristics, and computational fluid dynamic metrics, such as wall shear stress. An overlooked variable that intrinsically relates to intracranial atherosclerosis is the location of the arterial segment where the lesion occurs. Variations in cerebral blood flow, arterial anatomy, and flow dynamics are likely drivers of initial lesion development and thus likely to influence stroke recurrence. To date, treatment trials of intracranial atherosclerosis have not considered arterial segment lesion location as an independent variable, failing to account for variations in flow dynamics between each artery. There are limited available data on differences between arterial segments, confined to only post hoc analyses. In this review, we summarize available data on such differences between arterial segments. With the limited arterial segment data available, multiple differences in recurrence of stroke in territory of lesion were identified across trials.

HRU-Net: A high-resolution convolutional neural network for esophageal cancer radiotherapy target segmentation

Background and objectiveThe effective segmentation of esophageal squamous carcinoma lesions in CT scans is significant for auxiliary diagnosis and treatment. However, accurate lesion segmentation is still a challenging task due to the irregular form of the esophagus and small size, the inconsistency of spatio-temporal structure, and low contrast of esophagus and its peripheral tissues in medical images. The objective of this study is to improve the segmentation effect of esophageal squamous cell carcinoma lesions. MethodsIt is critical for a segmentation network to effectively extract 3D discriminative features to distinguish esophageal cancers from some visually closed adjacent esophageal tissues and organs. In this work, an efficient HRU-Net architecture (High-Resolution U-Net) was exploited for esophageal cancer and esophageal carcinoma segmentation in CT slices. Based on the idea of localization first and segmentation later, the HRU-Net locates the esophageal region before segmentation. In addition, an Resolution Fusion Module (RFM) was designed to integrate the information of adjacent resolution feature maps to obtain strong semantic information, as well as preserve the high-resolution features. ResultsCompared with the other five typical methods, the devised HRU-Net is capable of generating superior segmentation results. ConclusionsOur proposed HRU-NET improves the accuracy of segmentation for squamous esophageal cancer. Compared to other models, our model performs the best. The designed method may improve the efficiency of clinical diagnosis of esophageal squamous cell carcinoma lesions.

Segmentation of skin lesion using superpixel guided generative adversarial network with dual-stream patch-based discriminators

Accurate skin lesion segmentation in dermoscopic images is crucially important for the early diagnosis of skin cancer. Nevertheless, the existence of complex interfering factors, such as natural and artificial artifacts (e.g., hair and air bubbles), irregular appearance (e.g., varied shapes and low contrast), and significant differences in color shades cause the segmentation of skin lesion challenging. In this study, we propose a superpixel-guided generative adversarial network (GAN) with dual-stream patch-based discriminators for segmentation of the skin lesion. Specifically, in the designed GAN, a new multi-scale context extraction module (MCEM) is designed in the mask generator to enrich contextual information and capture boundary features of the lesion region, and thus accurately locating the object boundary of the lesion. Meantime, another branch of superpixel guided discriminator is designed and added into the discriminator module, which supplies so more compact and semantic information that enhances the discriminative ability. By using the dual-branch patch-based discriminators, the fine-grained discriminative power to local segmentation details is further enhanced and in turn making the generator produce more accurate segmentation masks. Comprehensive experiments show that our presented model achieves significant segmentation performance on the ISIC2016, ISIC2018, and HAM10000 skin lesion challenge datasets, and outperforms several promising deep convolutional neural networks.

Irregular adaptive refinement network for semantic segmentation of high-resolution remote sensing images

The semantic segmentation of high-resolution remote sensing images has broad application prospects in land cover classification, road extraction, urban planning and other fields. To alleviate the influence of the large data volume and complex background of high-resolution remote sensing images, the usual approach is to downsample them or cut them into small pieces for separate processing. Even if combining the two methods can improve the segmentation efficiency, it ignores the differences between the middle and the edge regions. Therefore, we consider the characteristics of large and irregular region in high-resolution remote sensing images, and then propose an irregular adaptive refinement network to locate the irregular edge region, which will be refined adaptively. Specifically, on the basis of effectively preserving the global and local information, the prediction confidence is calculated to locate pixel points that are poorly segmented, so as to form irregular regions requiring further refinement, avoiding to ‘over-refine’ intermediate region with good segmentation. At the same time, considering the difference in the refinement degree of different pixels, we propose to adaptively integrate the local segmentation results to refine the coarse segmentation results. In addition, in order to bridge the gap between the two extreme ends of the scale space, we introduce a multi-scale framework. Finally, we conducted experiments on the Deepglobe dataset showing that the proposed method performed 0.37% to 0.87% better than the previous state-of-the-art methods in terms of mean Intersection over Union (mIoU).

Exploring consumer segments for online food provisioning services: Online supermarkets vs. alternative food networks

Recent years have witnessed a surge in online food provisioning services spanning from online mass retail to alternative food networks with digital sales platforms. Hence, enhancing our understanding of consumer preferences in the digital food space is vital, as digitalization holds significant potential to alter consumer habits within the food sector. To address this, we surveyed a representative sample of Norwegian consumers. We employed best-worst scaling to assess their preferences for eighteen characteristics derived from a variety of online food provisioning services. The characteristics covered business model elements such as key partners, resources, activities, distribution channels, and customer relations. Our findings identify two consumer segments. The first segment, labeled the Convenience segment, prioritizes efficient delivery and user-friendly interfaces, characteristics that resonate with online supermarket services. The second segment labeled the Local segment, prioritizes a wide range of locally and sustainably produced food in line with the ethos of alternative food networks. Notably, the latter segment, accounting for 48 % of the respondents, views several of the convenience characteristics derived from online mass retail as important. Taking the convenience preferences of the Local segment into account could broaden the appeal of alternative food networks and other sustainability-oriented niche providers to a wider range of consumers.

Semantic prior guided fine-grained facial expression manipulation

AbstractFacial expression manipulation has gained wide attention and has been applied in various fields, such as film production, electronic games, and short videos. However, existing facial expression manipulation methods often overlook the details of local regions in images, resulting in the failure to preserve local structures and textures of images. To solve this problem, this paper proposes a local semantic segmentation mask-based GAN (LSGAN) to generate fine-grained facial expression images. LSGAN is composed of a semantic mask generator, an adversarial autoencoder, a transformative generator, and an AU-intensity discriminator. Our semantic mask generator generates eye, mouth, and cheek masks of face images. Then, our transformative generator integrates target expression labels and corresponding facial region features to generate a vivid target facial expression image. In this fashion, we can capture expressions from target face images explicitly. Furthermore, an AU-intensity discriminator is designed to capture facial expression variations and evaluate quality of generated images. Extensive experiments demonstrate that our method achieves authentic face images with accurate facial expressions and outperforms state-of-the-art methods qualitatively and quantitatively.

Deep Learning Framework for Liver Tumor Segmentation

INTRODUCTION: Segregating hepatic tumors from the liver in computed tomography (CT) scans is vital in hepatic surgery planning. Extracting liver tumors in CT images is complex due to the low contrast between the malignant and healthy tissues and the hazy boundaries in CT images. Moreover, manually detecting hepatic tumors from CT images is complicated, time-consuming, and needs clinical expertise.
 OBJECTIVES: An automated liver and hepatic malignancies segmentation is essential to improve surgery planning, therapy, and follow-up evaluation. Therefore, this study demonstrates the creation of an intuitive approach for segmenting tumors from the liver in CT scans.
 METHODS: The proposed framework uses residual UNet (ResUNet) architecture and local region-based segmentation. The algorithm begins by segmenting the liver, followed by malignancies within the liver envelope. First, ResUNet trained on labeled CT images predicts the coarse liver pixels. Further, the region-level segmentation helps determine the tumor and improves the overall segmentation map. The model is tested on a public 3D-IRCADb dataset.
 RESULTS: Two metrics, namely dice coefficient and volumetric overlap error (VOE), were used to evaluate the performance of the proposed method. ResUNet model achieved dice of 0.97 and 0.96 in segmenting liver and tumor, respectively. The value of VOE is also reduced to 1.90 and 0.615 for liver and tumor segmentation.
 CONCLUSION: The proposed ResUNet model performs better than existing methods in the literature. Since the proposed model is built using U-Net, the model ensures quality and precise dimensions of the output.

Advancements in acne detection: application of the CenterNet network in smart dermatology.

Acne detection is critical in dermatology, focusing on quality control of acne imagery, precise segmentation, and grading. Traditional research has been limited, typically concentrating on singular aspects of acne detection. We propose a multi-task acne detection method, employing a CenterNet-based training paradigm to develop an advanced detection system. This system collects acne images via smartphones and features multi-task capabilities for detecting image quality and identifying various acne types. It differentiates between noninflammatory acne, papules, pustules, nodules, and provides detailed delineation for cysts and post-acne scars. The implementation of this multi-task learning-based framework in clinical diagnostics demonstrated an 83% accuracy in lesion categorization, surpassing ResNet18 models by 12%. Furthermore, it achieved a 76% precision in lesion stratification, outperforming dermatologists by 16%. Our framework represents a advancement in acne detection, offering a comprehensive tool for classification, localization, counting, and precise segmentation. It not only enhances the accuracy of remote acne lesion identification by doctors but also clarifies grading logic and criteria, facilitating easier grading judgments.

Anatomical attention can help to segment the dilated pancreatic duct in abdominal CT

PurposePancreatic duct dilation is associated with an increased risk of pancreatic cancer, the most lethal malignancy with the lowest 5-year relative survival rate. Automatic segmentation of the dilated pancreatic duct from contrast-enhanced CT scans would facilitate early diagnosis. However, pancreatic duct segmentation poses challenges due to its small anatomical structure and poor contrast in abdominal CT. In this work, we investigate an anatomical attention strategy to address this issue.MethodsOur proposed anatomical attention strategy consists of two steps: pancreas localization and pancreatic duct segmentation. The coarse pancreatic mask segmentation is used to guide the fully convolutional networks (FCNs) to concentrate on the pancreas’ anatomy and disregard unnecessary features. We further apply a multi-scale aggregation scheme to leverage the information from different scales. Moreover, we integrate the tubular structure enhancement as an additional input channel of FCN.ResultsWe performed extensive experiments on 30 cases of contrast-enhanced abdominal CT volumes. To evaluate the pancreatic duct segmentation performance, we employed four measurements, including the Dice similarity coefficient (DSC), sensitivity, normalized surface distance, and 95 percentile Hausdorff distance. The average DSC achieves 55.7%, surpassing other pancreatic duct segmentation methods on single-phase CT scans only.ConclusionsWe proposed an anatomical attention-based strategy for the dilated pancreatic duct segmentation. Our proposed strategy significantly outperforms earlier approaches. The attention mechanism helps to focus on the pancreas region, while the enhancement of the tubular structure enables FCNs to capture the vessel-like structure. The proposed technique might be applied to other tube-like structure segmentation tasks within targeted anatomies.

Machine learning based endothelial cell image analysis of patients undergoing descemet membrane endothelial keratoplasty surgery.

In this study, we developed a machine learning approach for postoperative corneal endothelial cell images of patients who underwent Descemet's membrane keratoplasty (DMEK). An AlexNet model is proposed and validated throughout the study for endothelial cell segmentation and cell location determination. The 506 images of postoperative corneal endothelial cells were analyzed. Endothelial cell detection, segmentation, and determining of its polygonal structure were identified. The proposed model is based on the training of an R-CNN to locate endothelial cells. Next, by determining the ridges separating adjacent cells, the density and hexagonality rates of DMEK patients are calculated. The proposed method reached accuracy and F1 score rates of 86.15 % and 0.857, respectively, which indicates that it can reliably replace the manual detection of cellsin vivoconfocal microscopy (IVCM). The AUC score of 0.764 from the proposed segmentation method suggests a satisfactory outcome. A model focused on segmenting endothelial cells can be employed to assess the health of the endothelium in DMEK patients.

Advantages of laparoscopic segmentectomy of the liver using ICG fluorescent navigation by the negative staining method: A comparison with open procedure.

Laparoscopic segmentectomy (LS) using indocyanine green (ICG) fluorescence navigation with negative staining method has potential for performing accurate and safe anatomical excision. This study aimed to evaluate the significance of LS using ICG fluorescence navigation compared with open segmentectomy (OS). Eighty-seven patients who underwent anatomical segmentectomies were evaluated for OS (n = 44) and LS (n = 43). The Glissonean pedicle approach was performed using either extra- or intrahepatic method, depending on the location of segment in LS. After clamping pedicle, negative staining method was performed. Liver transection was done along intersegmental plane visualizing by overlay mode of ICG camera. Surgical outcomes were compared between two groups. Correlation between predicted resecting liver volume (PRLV) calculated using volumetry and actual resected liver volume (ARLV) was assessed in two groups. Patients who underwent LS showed better outcomes in operative time, blood loss, and length of hospital stay. There were significantly fewer Grade II and Grade III or higher postoperative complications in LS group. Both values of AST (p < 0.001) and ALT (p < 0.001) on postoperative day 1 were significantly lower in LS group than in OS group. PRLV and ARLV were more strongly correlated in LS (r = 0.896) than in OS (r = 0.773). The difference between PRLV and ARLV was significantly lower in LS group than in OS group (p = 0.022), and this trend was particularly noticeable in posterosuperior segment (p = 0.008) than in anterolateral segment (p = 0.811). LS using ICG navigation allows precise resection and may contribute to safer short-term outcomes than OS, particularly in posterosuperior segment.