Dynamics Trajectories Research Articles

A Dynamic Lane-Changing Trajectory Planning Algorithm for Intelligent Connected Vehicles Based on Modified Driving Risk Field Model

A dynamic LC trajectory planning algorithm based on the modified driving risk field is proposed to address the issue of dynamic changes during the lane-changing (LC) process. First, a modified driving risk field (MDRF) model is constructed for LC scenarios. Then, according to the state of the target vehicle and discrete sampling points, a series of LC candidate trajectories were generated based on the quintic polynomial. After eliminating candidate trajectories that do not meet the constraints, the MDRF was utilized as a safety evaluation function. Additionally, comfort and smoothness evaluation functions were combined to evaluate candidate LC trajectories in order to obtain the optimal LC reference trajectory. Then, this paper proposes a dynamic LC trajectory planning algorithm, addressing the challenges of complex traffic scenarios and dynamic changes in adjacent vehicle states. Utilizing the optimal reference trajectory as a basis, a dynamic segmented algorithm is applied to the x–t and y–x curves, constructing an optimized objective function that considers the MDRF. Under multiple constraints, including the continuity and smoothness of the lateral and longitudinal trajectories, the penalty function approach is employed to solve the optimization objective function, yielding the optimal LC trajectory adapted to real-time changes in the traffic state. Finally, the proposed dynamic LC trajectory planning algorithm was validated under four different scenarios by using MATLAB 2023b. The simulation results indicate the safety, continuity, and dynamic feasibility of the proposed algorithm. Moreover, it demonstrates strong adaptability and flexibility in challenging dynamic LC scenarios.

Colonial Echoes: Unraveling Economic Legacies and Geopolitical Shifts in the South Pacific Islands

Abstract The research delves into the intricate geopolitical dynamics underpinning the economy of the South Pacific region, focusing on attaining self-sufficiency and environmental sustainability amidst competing interests from global powers such as the United States, Australia, Japan, India, New Zealand, and China. It scrutinizes the geopolitical and geoeconomic factors influencing Pacific nations, curating data from academic papers, policy documents, media, and relevant websites. Challenges, opportunities, and concerns facing the region are discerned within historical and contemporary contexts, emphasizing the need for Pacific nations to integrate with China and India’s economies. Recommendations prioritize transparent, sustainable partnerships, advocating for alignment with Chinese and Indian economic prowess while safeguarding regional interests and values. A mixed-methods research framework is employed, upholding ethical standards and furnishing pragmatic recommendations for policymakers, international entities, and local stakeholders. Emphasizing reconciliation, community empowerment, and sustainable development, the research aims for a more equitable and resilient future. By illuminating the enduring imprint of colonialism on the region’s socio-economic trajectory and geopolitical dynamics, this study contributes to understanding and navigating the complexities of colonial legacies in the South Pacific Islands.

Unveiling Carbon Cluster Coating in Graphene CVD on MgO: Combining Machine Learning Force field and DFT Modeling.

In this study, we investigate the behavior of carbon clusters (Cn, where n ranges from 16 to 26) supported on the surface of MgO. We consider the impact of doping with common impurities (such as Si, Mn, Ca, Fe, and Al) that are typically found in ores. Our approach combines density functional theory calculations with machine learning force field molecular dynamics simulations. It is found that the C21 cluster, featuring a core-shell structure composed of three pentagons isolated by three hexagons, demonstrates exceptional stability on the MgO surface and behaves as an "enhanced binding agent" on MgO-doped surfaces. The molecular dynamics trajectories reveal that the stable C21 coating on the MgO surface exhibits less mobility compared to other sizes Cn clusters and the flexible graphene layer on MgO. Furthermore, this stability persists even at temperatures up to 1100K. The analysis of the electron localization function and potential function of Cn on MgO reveals the high localization electron density between the central carbon of the C21 ring and the MgO surface. This work proposes that the C21 island serves as a superstable and less mobile precursor coating on MgO surfaces. This explanation sheds light on the experimental defects observed in graphene products, which can be attributed to the reduced mobility of carbon islands on a substrate that remains frozen and unchanged.

Na Vacancy-Driven Phase Transformation and Fast Ion Conduction in W-Doped Na3SbS4 from Machine Learning Force Fields.

Solid-state sodium batteries require effective electrolytes that conduct at room temperature. The Na3PnCh4 (Pn = P, Sb; Ch = S, Se) family has been studied for their high Na ion conductivity. The population of Na vacancies, which mediate ion diffusion in these materials, can be enhanced through aliovalent doping on the pnictogen site. To probe the microscopic role of extrinsic doping and its impact on diffusion and phase stability, we trained a machine learning force field for Na3-x W x Sb1-x S4 based on an equivariant graph neural network. Analysis of large-scale molecular dynamics trajectories shows that an increased Na vacancy population stabilizes the global cubic phase at lower temperatures with enhanced Na ion diffusion and that the explicit role of the substitutional W dopants is limited. In the global cubic phase, we observe large and long-lived deviations of atoms from the averaged symmetry, echoing recent experimental suggestions. Evidence of correlated Na ion diffusion is also presented that underpins the suggested superionic nature of these materials.

Exploring Chemical Space to Identify Partial Binders Against hMPV Nucleocapsid Protein.

Human metapneumovirus (hMPV) has gained prominence in recent times as the predominant etiological agent of acute respiratory tract infections. This virus targets children, the elderly, and individuals with compromised immune systems. Given the protracted duration of hMPV transmission, it is probable that the majority of children will have acquired the virus by the age of 5. In individuals with compromised immune systems, recurrence of hMPV infection is possible. As hMPV matures, it remains latent from the time of acquisition. The genome of hMPV encompasses a pivotal protein referred to as the nucleocapsid protein (N). This protein assumes the form of a left-handed helical nucleocapsid, enveloping the viral RNA genome. The primary function of this structure is to protect nucleases, rendering it a potentially promising target for therapeutic advancements. The present study employs a methodology that involves structure-based virtual screening, followed by molecular dynamics simulation at a 250-ns time scale, to identify potential natural molecules or their derivatives from the ZINC Database. These molecules are investigated for their binding properties against the hMPV nucleoprotein. Based on an evaluation of the docking score, binding site interaction, and molecular dynamics studies, it has been found that two naturally occurring molecules, namely M1 (ZINC85629735) and M3 (ZINC85569125), have shown notable docking scores of -9.6and -10.7 kcal/mol, acceptable RMSD, RMSF, Rg, and soon calculated from molecular dynamics trajectory associated with MMGBSA binding energy of -81.94 and -99.63 kcal/mol, respectively. These molecules have shown the highest binding affinity towardnucleocapsid protein and demonstrated promising attributes as potential binders against hMPV.

Computing linear optical spectra in the presence of nonadiabatic effects on graphics processing units using molecular dynamics and tensor-network approaches.

We compare two recently developed strategies, implemented in open source software packages, for computing linear optical spectra in condensed phase environments in the presence of nonadiabatic effects. Both approaches rely on computing excitation energy and transition dipole fluctuations along molecular dynamics (MD) trajectories, treating molecular and environmental degrees of freedom on the same footing. Spectra are then generated in two ways: in the recently developed Gaussian non-Condon theory, the linear response functions are computed in terms of independent adiabatic excited states, with non-Condon effects described through spectral densities of transition dipole fluctuations. For strongly coupled excited states, we instead parameterize a linear vibronic coupling Hamiltonian directly from spectral densities of energy fluctuations and diabatic couplings computed along the MD trajectory. The optical spectrum is then calculated using powerful, numerically exact tensor-network approaches. Both the electronic structure calculations to sample system fluctuations and the quantum dynamics simulations using tensor-network methods are carried out on graphics processing units, enabling rapid calculations on complex condensed phase systems. We assess the performance of the approaches using model systems in the presence of a conical intersection and the pyrazine molecule in different solvent environments.

Neural dynamics of reversal learning in the prefrontal cortex and recurrent neural networks.

In probabilistic reversal learning, the choice option yielding reward at higher probability switches at a random trial. To perform optimally in this task, one has to accumulate evidence across trials to infer the probability that a reversal has occurred. In this study, we investigated how this reversal probability is represented in cortical neurons by analyzing the neural activity in prefrontal cortex of monkeys and recurrent neural networks trained on the task. We found that neural trajectories encoding reversal probability had substantial dynamics associated with intervening behaviors necessary to perform the task. Furthermore, the neural trajectories were translated systematically in response to whether outcomes were rewarded, and their position in the neural subspace captured information about reward outcomes. These findings suggested that separable dynamic trajectories, instead of fixed points on a line attractor, provided a better description of neural representation of reversal probability. Near the behavioral reversal, in particular, the trajectories shifted monotonically across trials with stable ordering, representing varying estimates of reversal probability around the reversal point. Perturbing the neural trajectory of trained networks biased when the reversal trial occurred, showing the role of reversal probability activity in decision-making. In sum, our study shows that cortical neurons encode reversal probability in a family of dynamic neural trajectories that accommodate flexible behavior while maintaining separability to represent distinct probabilistic values.

Dynamic modeling and trajectory optimization for the rigid-flexible coupled spacecraft with the free-floating manipulator and solar panels

The rigid-flexible coupled spacecraft, composed of flexible solar panels and a multilink manipulator, has gained prominence in on-orbit servicing due to rapid advancements in space technology. However, the intricate effects of rigid-flexible coupling pose significant challenges for dynamic modeling, trajectory planning, and control. This paper aims to develop general dynamic approaches for modeling and trajectory planning in such spacecraft, considering large deformations. The main distinguishing feature is the use of the referenced nodal coordinate formulation to accurately describe the large-deformed solar panels rather than directly treating them as disturbance for the free-floating system. Additionally, the common recursive model for the multilink manipulator is integrated into the same framework. The modal reduction method with modal derivatives techniques is employed to address geometric nonlinearity resulting from large deformations. Polynomial trajectory parameters with different performance characteristics are obtained by defining various optimization objectives. The coupling analysis is conducted based on an accurate reduced-order dynamic model, the results of which can be used for designing manipulator tasks. Coupling values are defined as the objective for trajectory optimization, offering advantages such as insensitivity to dynamic model accuracy and a fast optimization process. After validating the accuracy of the proposed dynamic model through simulations, the performances of different optimized trajectories are demonstrated using numerical examples.

A dosimetrically motivated pathfinding approach for non-isocentric dynamic trajectory radiotherapy

Objective.Non-isocentric dynamic trajectory radiotherapy (DTRT) involves dynamic table translations in synchrony with intensity modulation and dynamic gantry, table, and/or collimator rotation. This work aims to develop and evaluate a novel dosimetrically motivated path determination technique for non-isocentric DTRT.Approach.The path determination considers all available beam directions, given on a user-specified grid of gantry angle, table angle, and longitudinal, vertical, and lateral table position. Additionally, the source-to-target distance of all beam directions can be extended by moving the table away from the gantry along the central beam axis to increase the collision-free space. The path determination uses a column generation algorithm to iteratively add beam directions to paths until a user-defined total path length is reached. A subsequent direct aperture optimization of the intensity modulation along the paths creates deliverable plans. Non-isocentric DTRT plans using the path determination and using a manual path setup were created for a craniospinal and a spinal irradiation case. Furthermore, VMAT, isocentric DTRT, and non-isocentric DTRT plans are created for a breast, head and neck (H&N), and esophagus case. Additionally, a HyperArc plan is created for the H&N case. The plans are compared in terms of the dosimetric treatment plan quality and estimated delivery time.Main results.For the craniospinal and spinal irradiation case, using path determination results in dose distributions with improved conformity but a slightly worse target homogeneity compared to manual path setup. The non-isocentric DTRT plans maintained target coverage while reducing the mean dose to organs-at-risk on average by 1.7 Gy (breast), 1.0 Gy (H&N), and 1.6 Gy (esophagus) compared to the VMAT plans and by 0.8 Gy (breast), 0.6 Gy (H&N), and 0.8 Gy (esophagus) compared to the isocentric DTRT plans.Significance.A general dosimetrically motivated path determination applicable to non-isocentric DTRT plans is successfully developed, further advancing the treatment planning for non-isocentric DTRT.

DFCNet +: Cross-modal dynamic feature contrast net for continuous sign language recognition

In sign language communication, the combination of hand signs and facial expressions is used to convey messages in a fluid manner. Accurate interpretation relies heavily on understanding the context of these signs. Current methods, however, often focus on static images, missing the continuous flow and the story that unfolds through successive movements in sign language. To address this constraint, our research introduces the Dynamic Feature Contrast Net Plus (DFCNet+), a novel model that incorporates both dynamic feature extraction and cross-modal learning. The dynamic feature extraction module of DFCNet+ uses dynamic trajectory capture to monitor and record motion across frames and apply key features as an enhancement tool that highlights pixels that are critical for recognizing important sign language movements, allowing the model to follow the temporal variation of the signs. In the cross-modal learning module, we depart from the conventional approach of aligning video frames with textual descriptions. Instead, we adopt a gloss-level alignment, which provides a more detailed match between the visual signals and their corresponding text glosses, capturing the intricate relationship between what is seen and the associated text. The enhanced proficiency of DFCNet+ in discerning inter-frame details translates to heightened precision on benchmarks such as PHOENIX14, PHOENIX14-T and CSL-Daily. Such performance underscores its advantage in dynamic feature capture and inter-modal learning compared to conventional approaches to sign language interpretation. Our code is available at https://github.com/fyzjut/DFCNet_Plus.

Hydrogen Adsorption on Solid and Liquid Surfaces of Ga–Sn and Ga–In

AbstractDynamic liquid metal catalysts have remarkable activity and selectivity toward certain reactions, such as carbon dioxide reduction. The surface hydrogen coverage may play a role in this, including controlling competing reactions such as hydrogen evolution. Using a novel method of statistically selecting relevant snapshots from a dynamic liquid metal trajectory, the hydrogen adsorption energy is reported across liquid surfaces of Ga–Sn and Ga–In via density functional theory. A fully dynamic sampling of hydrogen adsorption to the liquid metal is also conducted with ab initio molecular dynamics at temperature. The results indicate that hydrogen only associates weakly with Ga–Sn and Ga–In surfaces, with minimal difference between the two materials. Hydrogen adsorbs only slightly more stably (≈0.1 eV) to the liquid surfaces compared to the solid. A low hydrogen coverage is predicted on the liquid metal of ≈0.03 H Å−2 at potentials of around –1.15 V (vs RHE). However, the mobility of hydrogen on the liquid surface is much greater due to a novel mechanism whereby the dynamic surface rearranges, opening pathways for diffusion. The results suggest that the unique catalytic behavior of these liquid metal materials may be due to changes in adsorbate diffusivity when the metals are melted.

Custom microfluidic chip design enables cost-effective three-dimensional spatiotemporal transcriptomics with a wide field of view.

Spatial transcriptomic techniques offer unprecedented insights into the molecular organization of complex tissues. However, integrating cost-effectiveness, high throughput, a wide field of view and compatibility with three-dimensional (3D) volumes has been challenging. Here we introduce microfluidics-assisted grid chips for spatial transcriptome sequencing (MAGIC-seq), a new method that combines carbodiimide chemistry, spatial combinatorial indexing and innovative microfluidics design. This technique allows sensitive and reproducible profiling of diverse tissue types, achieving an eightfold increase in throughput, minimal cost and reduced batch effects. MAGIC-seq breaks conventional microfluidics limits by enhancing barcoding efficiency and enables analysis of whole postnatal mouse sections, providing comprehensive cellular structure elucidation at near single-cell resolution, uncovering transcriptional variations and dynamic trajectories of mouse organogenesis. Our 3D transcriptomic atlas of the developing mouse brain, consisting of 93 sections, reveals the molecular and cellular landscape, serving as a valuable resource for neuroscience and developmental biology. Overall, MAGIC-seq is a high-throughput, cost-effective, large field of view and versatile method for spatial transcriptomic studies.

Meaning-oriented understanding of L2 academic writing development with task-type repetition

From a meaning-oriented perspective, the study drew on the nominalization in Systemic Functional Linguistics with Coh-Metrix and interviews to check how L2 academic writing of student writers may develop with the implementation of task-type repetition (TtR). The findings revealed that the development of L2 academic writing was presumably staged rather than constantly linear. Over the development, language proficiency dynamically resonated with TtR. First, the beginning session of TtR (from Task 1 to around Task 4) witnessed a more dynamic developmental trajectory during which the low-proficiency student constantly outperformed the other two in inter-sentential cohesion, while the intermediate-level student stood out over the middle session (Task 4- around Task 8). Second, at three genre stagings, students’ genre knowledge deeply resonated with their meaning-making activities throughout TtR, particularly in facilitating individual iterative reflection on their established writing knowledge. Third, the deficiency of genre knowledge may cause wide-range absence of nominalization at DESCRIPTION and CONCLUSION than EXPOSITION. Fourth, the linguistic fossilization identified at DESCRIPTION implies that anaphoric reconstrual may be individually mediated (by grammatically complicating the syntax in the study). Finally, a nuanced improvement in metaphorical enrichment and control was identified.

Deciphering the dynamic single-cell transcriptional landscape in the ocular surface ectoderm differentiation system

Abstract The ocular surface ectoderm (OSE) is essential for the development of the ocular surface, yet the molecular mechanisms driving its differentiation are not fully understood. In this study, we used single-cell transcriptomic analysis to explore the dynamic cellular trajectories and regulatory networks during the in vitro differentiation of embryonic stem cells (ESCs) into the OSE lineage. We identified nine distinct cell subpopulations undergoing differentiation along three main developmental branches: neural crest, neuroectodermal, and surface ectodermal lineages. Key marker gene expression, transcription factor activity, and signaling pathway insights revealed stepwise transitions from undifferentiated ESCs to fate-specified cell types, including a PAX6 + TP63 + population indicative of OSE precursors. Comparative analysis with mouse embryonic development confirmed the model’s accuracy in mimicking in vivo epiblast-to-surface ectoderm dynamics. By integrating temporal dynamics of transcription factor activation and cell–cell communication, we constructed a comprehensive molecular atlas of the differentiation pathway from ESCs to distinct ectodermal lineages. This study provides new insights into the cellular heterogeneity and regulatory mechanisms of OSE development, aiding the understanding of ocular surface biology and the design of cell-based therapies for ocular surface disorders.

Unveiling user interests: A deep user interest exploration network for sequential location recommendation

Sequential location recommendation, also known as next location prediction, is a crucial task aiming to compute the likelihood of a user visiting a certain point of interest (POI). It has significant applications in route planning and location-based advertisements. Despite the utilization of geographical information, existing methods often adopt a rigidly sequential prediction mode, focusing solely on predicting the next destination (i.e., where to go), which limits their ability to effectively model user trajectories in terms of predicting from where the user came. Moreover, these methods struggle to capture the evolving and deep user interests along their trajectory. To deal with these problems, we propose a novel approach called Deep User Interest Exploration Network (DUIEN) for sequential location recommendation. DUIEN incorporates two essential components to enhance recommendation accuracy. First, we employ a bidirectional neural network with the Cloze objective, enabling it to predict random masked POIs within user trajectories. This approach aids the model in gaining a better understanding of the context and dynamics of user trajectories. Second, we introduce a user interest exploration layer that seamlessly integrates into the sequential learning process. This layer effectively utilizes tag information associated with POIs and learns the intricate interplay between user interests and their trajectories using a deep user interest exploration method. In our experiments, DUIEN outperforms state-of-the-art sequential location recommendation models, demonstrating the effectiveness of the designed update mechanisms for deep user interest exploration. This showcases the potential of our approach to significantly enhance the accuracy of sequential location recommendation.

Distinctive features and prognostic utility of neutrophil in severe patients with Klebsiella pneumoniae infection.

Neutrophil plays a pivotal role in the management of Klebsiella pneumoniae infection. Delineate the clinical characteristics and prognostic utility of neutrophil in severe patients with K. pneumoniae infection are crucial for clinical management and prognostic assessment. K. pneumoniae patients with different infection sites were enrolled from Medical Information Mart for Intensive Care IV and eICU Collaborative Research Database. Temporal variations of neutrophil counts within 30 days of clinical onset were examined using locally weighted scatterplot smoothing curves. Logistic regression analysis was performed to assess the relationship between neutrophil counts and hospital mortality. A total of 1,705 patients caused by K. pneumonia were included in the study. The non-survivor group exhibited a comparatively older age and a higher proportion of K. pneumoniae infections originating from respiratory and bloodstream sources compared to the survivor group (38.4% vs 21.1%, p<0.0001, and 15.1% vs 10.3%, p=0.021). Patients combined with multiple drug resistance strains, respiratory infection, liver disease, and above 60 years exhibited a specific dynamic process of neutrophil levels. Neutrophils counts peaked at admission and 1-2 weeks later. There was a 'U'-shaped relationship between neutrophil counts and hospital mortality. Neutrophils in K. pneumoniae infected patients have distinctive features and dynamic clinical trajectories. Close monitoring of severe patients infected with K. pneumoniae upon admission and during the first 1-2 weeks after admission is of utmost importance, particularly for patients with a neutrophil count exceeding 8.0×109/L.

Trajectory of associative memory impairment during electroconvulsive therapy in depression

Memory impairment is a serious cognitive side effect of electroconvulsive therapy (ECT) in the treatment of major depressive episodes (MDEs) and has garnered widespread attention in clinical practice, but its underlying evolution pattern during the course of ECT remains rarely understood in detail. Associative memory (AM) is a core indicator that reflects memory impairment in ECT. This study aimed to identify the dynamic trajectory of AM impairment and explore associated predictive factors. 405 intensive longitudinal AM data from 81 patients with MDE were collected at the baseline, after the first, third, fifth, and eighth ECT using five sets of face-cued word memory paradigms. Changes in AM score over time were analyzed using a linear mixed effects model. Trajectory subgroups and predictive factors were investigated using growth mixture model and logistic regression. AM score during ECT were significantly lower than at baseline, with the lowest scores observed after the eighth ECT session. Two trajectories of rapid (N = 56, 69.14%) and slow (N = 25, 30.86%) AM impairment were differentiated. Older female with lower education level were significant predictors contributing to more rapid memory impairment for ECT. The evolving pattern of associative memory impairment during ECT appears to occur early and worsen with subsequent treatment. This study may provide the important evidence understanding of the number effect of ECT sessions on memory impairment and suggest individual factors for predicting ECT memory outcome.

A qualitative dynamic analysis of the relationship between tourism and human development

This study analyzes the dynamic relationship between tourism and human development in a sample of 123 countries between 1995–2019 using a symbolic time series methodological analysis, with the number of international tourist arrivals per capita as the tourism measurement variable and the Human Development Index as the development measurement variable. The objective was to determine if a higher level of tourism specialization is related to a higher level of economic development. The definition of economic regime is used and the concept of the distance between the dynamic trajectories of the different countries analyzed is introduced to create a minimum spanning tree. In this way, groups of countries are identified that display similar behavior in terms of tourism specialization and levels of human development. The results suggest that countries with a high level of tourism specialization have a higher level of development as compared to those in which tourism has a lower specific weight. However, the largest group of countries identified is characterized by low levels of tourism specialization and economic development, which appears to translate into a poverty trap. Therefore, policies related to tourism activity expansion should be created since higher tourism levels have been linked to higher levels of human development. In the case of less developed countries, however, these projects should be financed by international organizations so that these countries can escape the poverty trap in which they are currently found.

Reviewing approaches employed in Arabic chatbots

&lt;p&gt;The field of chatbots has witnessed a remarkable evolution in recent years, marked by a transition from simplistic rule-based structures to sophisticated systems employing advanced natural language processing (NLP) techniques. While most languages benefit from NLP support, the majority of chatbot research and development has been conducted in English, leaving a notable scarcity of comparable works in Arabic. This scarcity is attributed to the myriad challenges posed by the linguistically intricate nature of Arabic, encompassing orthographic variations and diverse dialects. This study systematically reviews articles that represent implementations of Arabic chatbots, revealing a discernible shift from rule-based frameworks to the predominant adoption of machine learning (ML) and deep learning (DL) methods. The results highlight the dynamic trajectory of chatbot technology, with a notable emphasis on the pivotal role of DL, as evidenced by a significant peak in 2023. Looking forward, the study anticipates a more sophisticated future for chatbot development, driven by ongoing advancements in artificial intelligence (AI) and NLP, offering valuable insights into the current state of Arabic chatbot research and laying the foundation for continued exploration in this evolving and dynamic field.&lt;/p&gt;

Subphenotypes of platelet count trajectories in sepsis from multi-center ICU data

Although thrombocytopenia on admission to the ICU is associated with increased in-hospital mortality in septic patients, the role of longitudinally measured platelet counts, which are dynamically changing, is unclear. We aimed to identify patterns of dynamic platelet count trajectories and evaluate their association with outcomes and thrombocytopenia in septic patients. We tested the longitudinal platelet trajectory patterns of sepsis patients within the first four days of ICU admission in the MIMIC-IV database and their association with 28-day mortality, and independently validated our findings in the eICU-CRD database. Statistical methods used included multivariate regression, propensity score analysis, doubly robust estimation, gradient boosting model, and inverse probability weighting to ensure the robustness of our findings. A total of 22,866 septic patients were included in our study. The trajectory analysis categorizes patients into ascending (AS), stable (ST), or descending (DS) patterns. The risk of 28-day mortality was increased in the DS patients (OR = 2.464, 95%CI 1.895–3.203, p < 0.001) and ST patients (OR = 1.302, 95%CI 1.067–1.589, p = 0.009) compared to AS patients. The AS patients had lower ICU length of stay (2.36 vs. 4.32, p < 0.001) and 28-day maximum SOFA scores (5.00 vs. 6.00, p < 0.001) than the DS patients, but had more ventilator-free days within 28 days than the DS group (26.00 vs. 24.00, p < 0.001). The mediating effect of thrombocytopenia was significant (p < 0.001 for the average causal mediation effect (ACME)). Longitudinal platelet trajectory was associated with risk-adjusted 28-day mortality among patients with sepsis and was proportionally mediated through thrombocytopenia.