Degeneracy‐Resistant LiDAR‐SLAM Algorithm Based on Geometric and Visual Features' Fusion

ABSTRACT Simultaneous localization and mapping (SLAM) is at the core of robotics automation, relying on sensors such as Laser Detection and Ranging (LiDAR) and cameras to digitally construct a robot's environment and determine its position within. LiDAR‐based SLAM outperforms visual‐SLAM, especially in low visibility and challenging lighting conditions. However, these systems still face challenges like scene degradation when dealing with feature‐deficient degenerate environments such as long corridors or tunnels. Traditional LiDAR SLAM algorithms primarily focus on the extraction of geometric features from the scene, with less utilization of visual information, for example, LiDAR‐generated reflectivity (also commonly referred to as intensity image) and depth imagery. In this study, we explore the potential of fusing both geometric and LiDAR‐generated image features into the SLAM system in various forms, aiming to enhance the system's adaptability in diverse environments and its robustness against environment degeneracy. We propose a new multifeature‐modality SLAM designed for robust real‐time localization and mapping in challenging environments. Our method enhances and extracts visual features from LiDAR‐generated images, which are then fused with geometric features through a holistic residual function for pose optimization. We also integrate a deep learning‐based object removal algorithm to reduce sensitivity to moving objects and sensor noise. This article conducts an in‐depth comparison of the proposed algorithm with several leading technologies in terms of scan matching accuracy, robustness, odometry, and mapping. The experimental results vividly showcase the superiority of our method in achieving high scan matching success rates and strong resilience against random outliers and Gaussian noise across various challenging scenarios, compared to the existing LiDAR SLAM methods that rely solely on geometric features. Extensive field experiments conducted on publicly available data sets, along with independently developed backpack‐based and robotic platforms, validated the robustness and accuracy of the proposed approach in both indoor and outdoor environments. In 3D mapping, we quantified the precision of 3D points by comparing point clouds collected by high‐precision Mobile Laser Scanning (MLS) and Terrestrial Laser Scanning (TLS). Our method outperforms in terms of absolute pose errors (APE) and point cloud matching quality. Based on the fitted Weibull distribution, the root mean square error (RMS) of point‐to‐plane distances improved by 20%. Additionally, ablation tests revealed the efficacy of different components within our system.

Numerical analysis of MHD rotating micropolar nanofluid flow characteristics using Buongiorno model

Abstract This computational study investigates the 3D flow, heat transfer, and mass transport properties of nanofluids, micropolar nanofluids, and Maxwell nanofluids over a stretching sheet in a rotating frame. This analysis employs the Buongiorno model, which incorporates binary chemical processes and Arrhenius activation energy. The governing equations are transformed via similarity variables and solved numerically with MATLAB's bvp4c solver. Validation against published data reveals outstanding consistency in velocity gradients and skin friction coefficients, with relative errors less than 1%. Results indicate that increasing the magnetic field( M ) $(M)$ and rotation( λ lambda)$ reduces velocities but increases temperature and nanoparticle concentration. The rmophoresis( N t ) $(Nt)$ and Brownian motion( N b ) $(Nb)$ elevate temperature while lowering the Nusselt number. The Maxwell nanofluid exhibits the slowest flow, the highest temperature, and concentration profiles, and lower Sherwood and Nusselt numbers compared to the other fluids under study. These results provide light on the unique thermal and hydrodynamic characteristics of complex nanofluids and have been confirmed against previous research. Optimizing nanofluid‐based thermal systems with rotation, magnetic fields, and non‐Newtonian effects—such as those used in rotating heat exchangers and microelectronic cooling—requires this knowledge.

Dissipativity-based robust output feedback control for finite-time boundedness of polytopic uncertain systems

A problem of designing an output feedback controller to improve transient performance in uncertain linear time-varying systems subjected to bounded external disturbances is considered. For systems with the dynamics specified by polytopic uncertainties, we design linear reduced-order dynamic controllers to bound the system state trajectories below a specified threshold over a predefined finite time interval. Using the notion of dissipativity, we also ensure that the system is robust with respect to bounded external disturbances over the finite time interval. Sufficient conditions for the uncertain closed-loop system trajectories to simultaneously satisfy finite-time boundedness and dissipativity to external disturbances are given in terms of parameter-dependent differential matrix inequalities. Further, parameter-independent differential linear matrix inequality (DLMI) conditions are presented to design time-varying control gains. The proposed dissipativity-based approach is less conservative as it generalises the procedure of designing finite-time robust controllers, encompassing various disturbance attenuation performance criteria like passivity and finite$ _2 $L 2-gain. The applicability of the proposed approach is demonstrated through numerical simulation examples.

ForestResU-Net: a hybrid deep learning model for low-resolution forest segmentation using digital orthophotos and RGB imagery

Accurate forest segmentation from aerial imagery is essential for forest inventory, ecological assessment, and sustainable management. However, low-resolution digital orthophotos (DOPs) and RGB imagery often present blurred boundaries, shadows, and mixed land cover, making precise segmentation challenging. This study aims to develop a robust deep learning approach capable of delivering reliable results under diverse geographical and environmental conditions. We propose a hybrid architecture, where additional residual blocks are incorporated into the encoder to enhance semantic feature extraction, while U-Net’s skip connections preserve spatial detail. The model was trained on a German dataset (GForest22) and tested on three unseen datasets: GForest23, GForest24 and DeepGlobe 2018. The proposed model achieved the highest average performance on GForest22 (Dice = 0.9329, IoU = 0.8775, Precision = 0.9831, Recall = 0.8912). Strong generalization was also demonstrated on DeepGlobe 2018, a cross-regional dataset, with Dice = 0.7918 and IoU = 0.7027. Across all datasets, the model consistently reduced false positives and maintained high precision while achieving competitive recall, even under complex land cover and shadow conditions. The findings confirm that the ForestResU-Net framework is scalable and transferable for large-scale forest segmentation from low-resolution imagery. The approach offers potential to support global forest monitoring, ecological evaluation, and sustainable environmental management.

Bioequivalence and pharmacokinetic prediction of oral marketed metformin extended-release tablets in Saudi Arabia

The Saudi market offers dosages of metformin that have varying release rates and can lead to different pharmacological reactions. Studying formulation bioequivalences helps patients choose the optimum medicine without affecting pharmacological responses. This study affects patient care clinically, which can improve Saudi diabetes management by improving efficacy, safety, adherence, access, and cultural relevance. The bioequivalence and pharmacokinetic parameters were studied using the convolution method. Similarity factor f 1 and different factors f 2 of different types of metformin tablets were calculated. Furthermore, the pharmacokinetic parameters were estimated using the convolution method. At pH = 6.8 (phosphate buffer) and 50 rpm, 100% of metformin was released within 2 h. While, after two hours in HCl at pH = 1.1 followed by five hours in phosphate buffer at pH = 6.8, 100% of the medication was released after 7 h. The release kinetics showed zero-order kinetics with r 2 = 0.961 for Formit ® and 0.971 for Glucophage ® , while the release mechanism showed that it follows the Higuchi equation with r 2 = 0.974 for Formit ® and 0.971 for Glucophage ® , respectively, indicating that the mechanism of drug release was controlled by diffusion. The two brands were lyo-equivalent, with a similarity factor and difference factor equal to 59.36 and 7.26, respectively. The convolution approach indicated that Glucophage ® and Formit ® have bioequivalent C max of 601 and 592 ng/ml, respectively. The two products had the same projected T max of 2.0 h and a modest AUC ∞ differential that did not violate the FDA’s 80-125% limit.

Real-world evidence of the impact of aspirin use on esophageal cancer incidence in Barrett’s esophagus patients

Barrett’s esophagus (BE) is the main precursor to esophageal adenocarcinoma, a cancer with a significantly rising incidence. While proton-pump inhibitors (PPIs) are the standard therapy for managing BE, the chemopreventive role of aspirin is an area of growing interest with inconclusive evidence, particularly regarding its use in combination with PPIs. This study aimed to assess whether adding aspirin to PPI therapy reduces the incidence of esophageal cancer in patients with BE more than PPIs alone. A nationwide retrospective cohort study was conducted using the TriNetX database. Adult patients with BE were divided into two cohorts: those receiving aspirin plus a PPI and those receiving a PPI only. Propensity score matching was used to balance baseline demographics and clinical comorbidities. The primary outcome was the incidence of malignant neoplasm of the lower third of the esophagus. Subgroup analyses were also performed for low-dose (81 mg) and high-dose (300–325 mg) aspirin. After matching, each cohort included 88,184 patients. The cohort receiving aspirin and PPIs had a lower risk of developing esophageal cancer compared to the PPI-only cohort (odds ratio [OR] 0.799, 95% CI: 0.679–0.941). The protective association was observed in both high-dose (OR 0.643) and low-dose (OR 0.664) aspirin subgroups, suggesting a potential dose-dependent effect. This large, real-world analysis suggests that the concurrent use of aspirin with PPIs is associated with a reduced risk of esophageal cancer in patients with BE.

Cerebral embolic protection devices in transcatheter aortic valve implantation: a systematic review and meta-analysis of randomized controlled trials

This meta-analysis aimed to assess the efficacy and safety of cerebral embolic protection devices (CEPDs) in patients undergoing transcatheter aortic valve implantation (TAVI). PubMed, Cochrane Central, and ScienceDirect were searched till April 2025. Risk ratios (RRs) with 95% confidence intervals (CIs) were pooled under a random-effects model using Review Manager. The Cochrane risk of bias (RoB 2.0) tool was used for quality assessment. Funnel plots were assessed for publication bias. Eight randomized controlled trials, including 11,632 patients undergoing TAVI, were analyzed. Use of CEPDs showed a non-significant reduction in all strokes (RR 0.92, 95% CI: 0.74–1.15, p  = 0.48) and disabling stroke (RR 0.80, 95% CI: 0.57–1.12, p  = 0.18). There was no significant difference in all-cause mortality (RR 1.09, 95% CI: 0.71–1.67, p  = 0.70), acute kidney injury (AKI) (RR 0.96, 95% CI: 0.44–2.11, p  = 0.93), disabling bleeding (RR 0.96, 95% CI: 0.28–3.31; p  = 0.94) and major vascular complications (RR 1.25, 95% CI: 0.56–2.78, p  = 0.59). CEPD did not significantly reduce the incidence of ischemic lesions or neurocognitive decline. Current evidence does not support a statistically significant clinical benefit of CEPD use during TAVI. While trends suggest a potential reduction in stroke, larger trials are needed to establish the significance of these results.

Heat transport analysis in buoyancy-driven flow of Maxwell fluid induced by a vertically stretching sheet inspired by Cattaneo Christov theory

We used non-Fourier's approach to model the heat transfer equations in magneto-hydrodynamic Maxwell nanofluid flow due to vertical stretching sheet under the influence of buoyancy force. The impact of uniform transverse magnetic field is also considered. This problem is analyzed with two heating processes: constant wall temperature (CWT) and prescribed surface temperature (PST). The heat and mass transport features are explored by implementing the modified Fourier's and Fick's laws.The governing equations are transformed into ordinary differential equations (ODEs) through appropriate similarity transformations. The resulting equations are solved numerically using the bvp4c program in MATLAB. The impact of physical parameter on the velocity, thermal, and solutal distributions is discussed in detail and presented graphically for aiding the flow. In this study, the velocity boosts unsteadiness parameter and thermal buoyancy parameter. Both velocity and temperature behave in opposite ways when affected by a magnetic field. Moreover, the building strength of the thermal relaxation time parameter does not escalate the thermal transport.

Institutional drivers of newborn hepatitis B vaccine disparities: A multi-hospital analysis in Washington, DC, 2017–2023

Despite universal newborn hepatitis B vaccination recommendations, birth dose coverage remains suboptimal with persistent racial disparities. While individual-level factors are well-studied, institutional practices’ role in vaccination outcomes remains poorly understood. We conducted a retrospective cohort study of 87,246 singleton births across eight Washington, DC hospitals from 2017–2023, using multilevel mixed-effects logistic regression to examine institutional disparities in vaccine refusal. Overall refusal rate was 6.7%, declining from 12.1% in 2017 to 3.5% in 2023. Hospital-specific rates varied dramatically, from near-zero to over 50%. Multilevel analysis revealed 31.1% of refusal variance was attributable to between-hospital and between-year differences, with stable hospital characteristics accounting for 71% of this contextual variance (22.1% of total) and temporal trends accounting for 29% (9.0% of total). Analysis of hospital-specific temporal trajectories revealed marked heterogeneity in response speed to the 2018 ACIP policy change: safety-net hospitals achieved target refusal rates within one year, while institutions with higher baseline refusal required 2–4 y. In contrast to national adult vaccination patterns where White individuals have higher coverage, White infants had lower refusal odds than Black infants after covariate adjustment, suggesting institutional practices may be associated with context-specific disparities that differ from broader population patterns. Sensitivity analyses confirmed robustness of findings. Substantial institutional variation in vaccination practices is associated with disparities beyond what would be expected from patient demographics alone, highlighting the need for system-level interventions targeting organizational factors to achieve equitable vaccination coverage.

In addition to ‘the evolution of predatory Journals - New strategies and threats. A letter to the editor’