Hybrid Model Research Articles

Intertumoral Heterogeneity Based on MRI Radiomic Features Estimates Recurrence in Hepatocellular Carcinoma.

Hepatocellular carcinoma (HCC) heterogeneity impacts prognosis, and imaging is a potential indicator. To characterize HCC image subtypes in MRI and correlate subtypes with recurrence. Retrospective. A total of 440 patients (training cohort = 213, internal test cohort = 140, external test cohort = 87) from three centers. 1.5-T/3.0-T, fast/turbo spin-echo T2-weighted, spin-echo echo-planar diffusion-weighted, contrast-enhanced three-dimensional gradient-recalled-echo T1-weighted with extracellular agents (Gd-DTPA, Gd-DTPA-BMA, and Gd-BOPTA). Three-dimensional volume-of-interest of HCC was contoured on portal venous phase, then coregistered with precontrast and late arterial phases. Subtypes were identified using non-negative matrix factorization by analyzing radiomics features from volume-of-interests, and correlated with recurrence. Clinical (demographic and laboratory data), pathological, and radiologic features were compared across subtypes. Among clinical, radiologic features and subtypes, variables with variance inflation factor above 10 were excluded. Variables (P < 0.10) in univariate Cox regression were included in stepwise multivariate analysis. Three recurrence estimation models were built: clinical-radiologic model, subtype model, hybrid model integrating clinical-radiologic characteristics, and subtypes. Mann-Whitney U test, Kruskal-Wallis H test, chi-square test, Fisher's exact test, Kaplan-Meier curves, log-rank test, concordance index (C-index). Significance level: P < 0.05. Two subtypes were identified across three cohorts (subtype 1:subtype 2 of 86:127, 60:80, and 36:51, respectively). Subtype 1 showed higher microvascular invasion (MVI)-positive rates (53%-57% vs. 26%-31%), and worse recurrence-free survival. Hazard ratio (HR) for the subtype is 6.10 in subtype model. Clinical-radiologic model included alpha-fetoprotein (HR: 3.01), macrovascular invasion (HR: 2.32), nonsmooth tumor margin (HR: 1.81), rim enhancement (HR: 3.13), and intratumoral artery (HR: 2.21). Hybrid model included alpha-fetoprotein (HR: 2.70), nonsmooth tumor margin (HR: 1.51), rim enhancement (HR: 3.25), and subtypes (HR: 5.34). Subtype model was comparable to clinical-radiologic model (C-index: 0.71-0.73 vs. 0.71-0.73), but hybrid model outperformed both (C-index: 0.77-0.79). MRI radiomics-based clustering identified two HCC subtypes with distinct MVI status and recurrence-free survival. Hybrid model showed superior capability to estimate recurrence. 3 TECHNICAL EFFICACY STAGE: 2.

Accurate kinematic calibration of a six-DoF serial robot by using hybrid models with reduced dimension and minimized linearization errors

Accurate kinematic calibration of a six-DoF serial robot by using hybrid models with reduced dimension and minimized linearization errors

Interpretable and explainable hybrid model for daily streamflow prediction based on multi-factor drivers.

Streamflow time series data typically exhibit nonlinear and nonstationary characteristics that complicate precise estimation. Recently, multifactorial machine learning (ML) models have been developed to enhance the performance of streamflow predictions. However, the lack of interpretability within these ML models raises concerns about their inner workings and reliability. This paper introduces an innovative hybrid architecture, the TCN-LSTM-Multihead-Attention model, which combines two layers of temporal convolutional networks (TCN) followed by one layer of long short-term memory (LSTM) units, integrated with a Multihead-Attention mechanism for predicting streamflow with streamflow causation-driven prediction samples (RCDP), employing local and global interpretability studies through Shapley values and partial dependency analysis. The find_peaks method was used to identify peak flow events in the test dataset, validating the model's generality and uncovering the physical causative patterns of streamflow. The results show that (1) compared to the LSTM model with the same hyperparameter settings, the proposed TCN-LSTM-Multihead-Attention hybrid model increased the R2 by 52.9%, 2.5%, 43.1%, and 10.7% respectively at four stations in the test set predictions using RCDP samples. Moreover, comparing the prediction results of the hybrid model under different samples in Hengshan station, the R2 for RCDP increased by 5.06% and 1.22% compared to streamflow autoregressive prediction samples (RAP) and meteorological-soil volumetric water content coupled autoregressive prediction samples (MCSAP) respectively. (2) Historical streamflow data from the preceding 3days predominantly influences predictions due to strong autocorrelation, with flow quantity (Q) typically emerging as the most significant feature alongside precipitation (P), surface soil moisture (SSM), and adjacent station flow data. (3) During periods of low and normal flow, historical data remains the most crucial factor; however, during flood periods, the roles of upstream inflow and precipitation become significantly more pronounced. This model facilitates the identification and quantification of various hydrodynamic impacts on flow predictions, including upstream flood propagation, precipitation, and soil moisture conditions. It also elucidates the model's nonlinear relationships and threshold responses, thereby enhancing the interpretability and reliability of streamflow predictions.

Pangenome and multi-tissue gene atlas provide new insights into the domestication and highland adaptation of yaks

BackgroundThe genetic diversity of yak, a key domestic animal on the Qinghai-Tibetan Plateau (QTP), is a vital resource for domestication and breeding efforts. This study presents the first yak pangenome obtained through the de novo assembly of 16 yak genomes.ResultsWe discovered 290 Mb of nonreference sequences and 504 new genes. Our pangenome-wide presence and absence variation (PAV) analysis revealed 5,120 PAV-related genes, highlighting a wide range of variety-specific genes and genes with varying frequencies across yak populations. Principal component analysis (PCA) based on binary gene PAV data classified yaks into three new groups: wild, domestic, and Jinchuan. Moreover, we proposed a ‘two-haplotype genomic hybridization model’ for understanding the hybridization patterns among breeds by integrating gene frequency, heterozygosity, and gene PAV data. A gene PAV-GWAS identified a novel gene (BosGru3G009179) that may be associated with the multirib trait in Jinchuan yaks. Furthermore, an integrated transcriptome and pangenome analysis highlighted the significant differences in the expression of core genes and the mutational burden of differentially expressed genes between yaks from high and low altitudes. Transcriptome analysis across multiple species revealed that yaks have the most unique differentially expressed mRNAs and lncRNAs (between high- and low-altitude regions), especially in the heart and lungs, when comparing high- and low-altitude adaptations.ConclusionsThe yak pangenome offers a comprehensive resource and new insights for functional genomic studies, supporting future biological research and breeding strategies.

Enhancing Coronary Artery Disease Detection with a Hybrid Machine Learning Approach: Integrating K-Nearest Neighbor (KNN) and Support Vector Machine (SVM) Algorithms

Recent studies have identified coronary artery disease (CAD) as a leading cause of death globally. Early detection of CAD is crucial for reducing mortality rates. However, accurately predicting CAD poses challenges, particularly in treating patients effectively before a heart attack occurs due to the complexity of data and relationships in traditional methodologies. This research has successfully developed a machine learning model for CAD prediction by combining K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) Classifier techniques. The model, trained and tested on a dataset of 918 samples (508 with cardiac issues and 410 healthy cases), achieved an accuracy of 82% for KNN, 84.3% for SVM, and 88.7% for the hybrid model after rigorous training and testing.

A cyber defense system against phishing attacks with deep learning game theory and LSTM-CNN with African vulture optimization algorithm (AVOA)

AbstractPhishing attacks pose a significant threat to online security, utilizing fake websites to steal sensitive user information. Deep learning techniques, particularly convolutional neural networks (CNNs), have emerged as promising tools for detecting phishing attacks. However, traditional CNN-based image classification methods face limitations in effectively identifying fake pages. To address this challenge, we propose an image-based coding approach for detecting phishing attacks using a CNN-LSTM hybrid model. This approach combines SMOTE, an enhanced GAN based on the Autoencoder network, and swarm intelligence algorithms to balance the dataset, select informative features, and generate grayscale images. Experiments on three benchmark datasets demonstrate that the proposed method achieves superior accuracy, precision, and sensitivity compared to other techniques, effectively identifying phishing attacks and enhancing online security.

Hybrid Transfer Learning and Support Vector Machine Models for Asphalt Pavement Distress Classification

Pavement condition evaluation plays a crucial role in assisting with the management of the highway infrastructure. However, the current methods used for assessing pavement conditions are costly, time-consuming, and subjective. There is a growing need to automate these assessment tactics and leverage low-cost technologies to enable widespread deployment. This study aims to develop robust and highly accurate models for classifying asphalt pavement distresses using transfer learning (TL) techniques based on pretrained deep learning (DL) networks. This topic has gained considerable attention in the field since 2015 when DL became the mainstream choice for various computer vision tasks. While progress has been made in TL model development, challenges persist in areas of accuracy, repeatability, and training cost. To tackle these challenges, this study proposes hybrid models that combine DL networks with support vector machines (SVMs). Three strategies were evaluated: single DL models using transfer learning (TLDL), hybrid models combining DL and SVM (DL+SVM), and hybrid models combining TLDL and SVM (TLDL+SVM). The performance of each strategy was assessed using statistical metrics based on the confusion matrix. Results consistently showed that the TLDL+SVM strategy outperformed the other approaches in accuracy and F1 scores, regardless of the DL network type. On average, the hybrid models achieved an accuracy of 95%, surpassing the 80% accuracy of the best single model and the 55% accuracy for DL+SVM without TL. The results clearly indicate that employing transfer-learned models as feature extractors, in combination with SVM as the classifier, consistently achieves exceptional performance.

Hybrid modeling of an ultracentrifugation process for separation of full and empty adeno-associated virus particles.

Ultracentrifugation is an attractive method for separating full and empty capsids, exploiting their density difference. Changes of the serotype/capsid, density of loading material, or the genetic information contained in the adeno-associated viruses (AAVs) require the adaptation of the harvesting parameters and the density gradient loaded onto the centrifuge. To streamline these adaptations, a mathematical model could support the design and testing of operating conditions.Here, hybrid models, which combine empirical functions with artificial neural networks, are proposed to describe the separation of full and empty capsids as a function of material and operational parameters, i.e., the harvest model. In addition, critical quality attributes are estimated by a quality model which is operating on top of the harvest model. The performance of these models was evaluated using test data and two additional blind runs. Also, a "what-if" analysis was conducted to investigate whether the models' predictions align with expectations.It is concluded that the models are sufficiently accurate to support the design of operating conditions, though the accuracy and applicability of the models can further be increased by training them on more specific data with higher variability.

Development of Hybrid Models Based on AlexNet and Machine Learning Approaches for Strip Steel Surface Defect Classification

Development of Hybrid Models Based on AlexNet and Machine Learning Approaches for Strip Steel Surface Defect Classification

Crystal Phase Ionic Liquids for Energy Applications: Heat Capacity Prediction via a Hybrid Group Contribution Approach

In the selection and design of ionic liquids (ILs) for various applications, including heat transfer fluids, thermal energy storage materials, fuel cells, and solvents for chemical processes, heat capacity is a key thermodynamic property. While several attempts have been made to develop predictive models for the estimation of the heat capacity of ILs in their liquid phase, none so far have been reported for the ILs’ solid crystal phase. This is particularly important for applications where ILs will be used for thermal energy storage in the solid phase. For the first time, a model has been developed and used for the prediction of crystal phase heat capacity based on extending and modifying a previously developed hybrid group contribution model (GCM) for liquid phase heat capacity. A comprehensive database of over 5000 data points with 71 unique crystal phase ILs, comprising 42 different cations and 23 different anions, was used for parameterization and testing. This hybrid model takes into account the effect of the anion core, cation core, and subgroups within cations and anions, in addition to the derived indirect parameters that reflect the effects of branching and distribution around the core of the IL. According to the results, the developed GCM can reliably predict the crystal phase heat capacity with a mean absolute percentage error of 6.78%. This study aims to fill this current gap in the literature and to enable the design of ILs for thermal energy storage and other solid phase applications.

Psychometric evaluation of the PCL-5: assessing validity, diagnostic utility, and bifactor structures

ABSTRACT Background: The changes DSM-5 brought to the diagnostic criteria for posttraumatic stress disorder (PTSD) resulted in revising the most widely used instrument in assessing PTSD, namely the Posttraumatic Checklist for DSM-5 (PCL-5). Objective: This study examined the psychometric properties of the Romanian version of the PCL-5, tested its diagnostic utility against the Structured Clinical Interview for DSM-5 (SCID-5), and investigated the latent structure of PTSD symptoms through correlated symptom models and bifactor modelling. Method: A total sample of 727 participants was used to test the psychometric properties and underlying structure of the PCL-5 and 101 individuals underwent clinical interviews using SCID-5. Receiver operating characteristic curve (ROC) analyses were performed to test the diagnostic utility of the PCL-5 and identify optimal cut-off scores based on Youden's J index. Confirmatory Factor Analyses (CFAs) and bifactor modelling were performed to investigate the latent structure of PTSD symptoms. Results: Estimates revealed that the PCL-5 is a valuable tool with acceptable diagnostic accuracy compared to SCID-5 diagnoses, indicating a cut-off score of >47. The CFAs provide empirical support for Anhedonia, Hybrid, and bifactor models. The findings are limited by using retrospective, self-report data and the high percentage of female participants. Conclusions: The PCL-5 is a psychometrically sound instrument that can be useful in making provisional diagnoses within community samples and improving trauma-informed practices.

Quantum Machine Learning for Credit Scoring

This study investigates the integration of quantum circuits with classical neural networks for enhancing credit scoring for small- and medium-sized enterprises (SMEs). We introduce a hybrid quantum–classical model, focusing on the synergy between quantum and classical rather than comparing the performance of separate quantum and classical models. Our model incorporates a quantum layer into a traditional neural network, achieving notable reductions in training time. We apply this innovative framework to a binary classification task with a proprietary real-world classical credit default dataset for SMEs in Singapore. The results indicate that our hybrid model achieves efficient training, requiring significantly fewer epochs (350) compared to its classical counterpart (3500) for a similar predictive accuracy. However, we observed a decrease in performance when expanding the model beyond 12 qubits or when adding additional quantum classifier blocks. This paper also considers practical challenges faced when deploying such models on quantum simulators and actual quantum computers. Overall, our quantum–classical hybrid model for credit scoring reveals its potential in industry, despite encountering certain scalability limitations and practical challenges.

Guidelines for fracture risk assessment and management of osteoporosis in postmenopausal women and men above the age of 50 in Qatar

SummaryWe present comprehensive guidelines for osteoporosis management in Qatar. Formulated by the Qatar Osteoporosis Association, the guidelines recommend the age-dependent Qatar fracture risk assessment tool for screening, emphasizing risk-based treatment strategies and discouraging routine dual-energy X-ray scans. They offer a vital resource for physicians managing osteoporosis and fragility fractures nationwide.PurposeOsteoporosis and related fragility fractures are a growing public health issue with an impact on individuals and the healthcare system. We aimed to present guidelines providing unified guidance to all healthcare professionals in Qatar regarding the management of osteoporosis.MethodsThe Qatar Osteoporosis Association formulated guidelines for the diagnosis and management of osteoporosis in postmenopausal women and men above the age of 50. A panel of six local rheumatologists who are experts in the field of osteoporosis met together and conducted an extensive review of published articles and local and international guidelines to formulate guidance for the screening and management of postmenopausal women and men older than 50 years in Qatar.ResultsThe guidelines emphasize the use of the age-dependent hybrid model of the Qatar fracture risk assessment tool for screening osteoporosis and risk categorization. The guidelines include screening, risk stratification, investigations, treatment, and monitoring of patients with osteoporosis. The use of a dual-energy X-ray absorptiometry scan without any risk factors is discouraged. Treatment options are recommended based on risk stratification.ConclusionGuidance is provided to all physicians across the country who are involved in the care of patients with osteoporosis and fragility fractures.

ICEEMDAN-Informer-GWO: a hybrid model for accurate wind speed prediction.

Extensive research has been diligently conducted on wind energy technologies in response to pressing global environmental challenges and the growing demand for energy. Accurate wind speed predictions are crucial for the effective integration of large wind power systems. This study presents a novel and hybrid framework called ICEEMDAN-Informer-GWO, which combines three components to enhance the accuracy of wind speed predictions. The improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) component improves the decomposition of wind speed data, the Informer model provides computationally efficient wind speed predictions, and the grey wolf optimisation (GWO) algorithm optimises the parameters of the Informer model to achieve superior performance. Three different sets of wind speed prediction (WSP) models and wind farm data from Block Island, Gulf Coast, and Garden City are used to thoroughly assess the proposed hybrid framework. This evaluation focusses on WSP for three specific time horizons: 5minutes, 30minutes, and 1hour ahead. The results obtained from the three conducted experiments conclusively demonstrate that the proposed hybrid framework exhibits superior performance, leading to statistically significant improvements across all three time horizons.

Phenological mismatch between trees and wildflowers: Reconciling divergent findings in two recent analyses

Abstract Recent evidence suggests that community science and herbarium datasets yield similar estimates of species' phenological sensitivities to temperature. Despite this, two recent studies by Alecrim et al. (2023) and Miller et al. (2022) found very different results when using different data sources (community science and herbarium specimens, respectively) to investigate whether warming threatens wildflowers with phenological mismatch in relation to shading by deciduous trees. Here, we investigated whether differences between the two studies' results could be reconciled by testing four hypotheses related to model design, species, spatiotemporal data extent and phenophase. Hybrid model structures brought results from the two datasets closer together but did not fully reconcile the differences between the studies. Neither the species nor the phenophase selected for analysis seemed to be responsible for differences in results. Cropping the datasets to match spatial and temporal extents appeared to reconcile most differences but only at the cost of much higher uncertainty associated with reduced sample size. Synthesis: Our analysis suggests that although species‐level estimates of phenological sensitivity may be similar between community science and herbarium datasets, inherent differences in the types and extent of data may lead to contradictory inference about complex biotic interactions. We conclude that, until community science data repositories expand to match the range of climate conditions present in herbarium collections or until herbarium collections match the spatial extent and temporal frequency of community science repositories, ecological studies should ideally be evaluated using both datasets to test the possibility of biased results from either.

Quantum deep learning-based anomaly detection for enhanced network security

Identifying and mitigating aberrant activities within the network traffic is important to prevent adverse consequences caused by cyber security incidents, which have been increasing significantly in recent times. Existing research mainly focuses on classical machine learning and deep learning-based approaches for detecting such attacks. However, exploiting the power of quantum deep learning to process complex correlation of features for anomaly detection is not well explored. Hence, in this paper, we investigate quantum machine learning and quantum deep learning-based anomaly detection methodologies to accurately detect network attacks. In particular, we propose three novel quantum auto-encoder-based anomaly detection frameworks. Our primary aim is to create hybrid models that leverage the strengths of both quantum and deep learning methodologies for efficient anomaly recognition. The three frameworks are formed by integrating the quantum autoencoder with a quantum one-class support vector machine, a quantum random forest, and a quantum k-nearest neighbor approach. The anomaly detection capability of the frameworks is evaluated using benchmark datasets comprising computer and Internet of Things network flows. Our evaluation demonstrates that all three frameworks have a high potential to detect the network traffic anomalies accurately, while the framework that integrates the quantum autoencoder with the quantum k-nearest neighbor yields the highest accuracy. This demonstrates the promising potential for the development of quantum frameworks for anomaly detection, underscoring their relevance for future advancements in network security.

Decomposition and Holt-Winters Enhanced by the Whale Optimization Algorithm for Forecasting the Amount of Water Inflow into the Large Dam Reservoirs in Southern Thailand

This study introduces hybrid forecasting models integrating the Whale Optimization Algorithm (WOA) with Holt-Winters (HW) and decomposition methods, applied in both additive and multiplicative models, for time series forecasting. Focusing on monthly water inflow into four dam reservoirs in Southern Thailand, the study compares these hybrid models against classical statistical models, Grid Search for Holt-Winters (Grid-HW) and Classical Decomposition (Classic-D). The analysis comprises two phases: the training dataset phase and the testing dataset phase. In the training phase, WOA demonstrates superior parameter optimization, enhancing both HW and decomposition methods, resulting in lower Mean Absolute Error (MAE) values compared to classical models. In the testing phase, performance metrics such as Root Mean Square Error (RMSE), MAE, and Symmetric Mean Absolute Percentage Error (sMAPE) are employed. The findings reveal that the Whale Optimization Algorithm with Holt-Winters (WOA-HW) and Decomposition (WOA-D) models surpass classical approaches in long-term forecasting accuracy for three dam reservoirs. Over 24 data points, the WOA with Multiplicative Holt-Winters (WOA-HWx) is optimal for Pran Buri dam, the WOA with Additive Decomposition (WOA-D+) for Bang Lang dam, and the WOA with Multiplicative Decomposition (WOA-Dx) for Kaeng Krachan dam. The Box-Jenkins approach, further refined through a Box-Cox transformation employing a natural logarithm, emerged as the superior forecasting model for Rajjaprabha dam. This model satisfied all critical statistical criteria, including normality of residuals (Anderson-Darling: 0.359, p-value: 0.433), homoscedasticity (Levene's test: 1.24, p-value: 0.274), independence (Ljung-Box test: 14.10, p-value: 0.169), and zero mean (t-test: -0.39, p-value: 0.702), establishing its robustness and reliability for forecast analysis.

Local Assessment and Small Bowel Crohn’s Disease Severity Scoring using AI

We present a machine learning and computer vision approach for a localized, automated, and standardized scoring of Crohn's disease (CD) severity in the small bowel, overcoming the current limitations of manual measurements CT enterography (CTE) imaging and qualitative assessments, while also considering the complex anatomy and distribution of the disease. Two radiologists introduced a severity score and evaluated disease severity at 7.5mm intervals along the curved planar reconstruction of the distal and terminal ileum using 236 CTE scans. A hybrid model, combining deep-learning, 3-D CNN, and Random Forest model, was developed to classify disease severity at each mini-segment. Precision, sensitivity, weighted Cohen's score, and accuracy were evaluated on a 20% hold-out test set. The hybrid model achieved precision and sensitivity ranging from 42.4% to 84.1% for various severity categories (normal, mild, moderate, and severe) on the test set. The model's Cohen's score (κ=0.83) and accuracy (70.7%) were comparable to the inter-observer agreement between experienced radiologists (κ=0.87, accuracy = 76.3%). The model accurately predicted disease length, correlated with radiologist-reported disease length (r=0.83), and accurately identified the portion of total ileum containing moderate-to-severe disease with an accuracy of 91.51%. The proposed automated hybrid model offers a standardized, reproducible, and quantitative local assessment of small bowel CD severity and demonstrates its value in CD severity assessment.

A hybrid model based on unsupervised learning for seismic response feature extraction and gas-bearing reservoir prediction using longitudinal and transverse waves seismic data in low-degree exploration areas: A case study

Predicting gas-bearing reservoirs from seismic data is a major field of research. Recently, deep learning has been applied to achieve better results; however, challenges such as high sample demand remain. Drilling data are often insufficient, and gas-bearing sample data are difficult to acquire in large quantities, particularly for low-level (new) exploration areas, which limits gas-bearing reservoir prediction by supervised learning. Therefore, this study proposes a hybrid model using unsupervised learning to extract seismic response features and predict gas-bearing reservoir distribution. First, seismic attribute analysis was used to extract as many longitudinal and transverse waves seismic attributes as possible to provide richer information about gas-bearing reservoirs. However, abundant seismic attributes contain a large amount of redundant information, thereby increasing computation time. Thus, principal component analysis (PCA) was used to reduce the dimensions of the extracted multi-component seismic attributes, remove redundant information, and obtain the most important characteristics that are sensitive to gas-bearing reservoirs. The sensitive seismic attributes obtained after the PCA were fed into a self-organizing neural network (SOM) for unsupervised learning. Neurons in the SOM compete with each other to optimize their own features and extract reservoir-sensitive seismic response features for prediction. To evaluate the proposed method, it was applied to part of the Marmousi2 model and real seismic data for validation. The following results were obtained: (1)The results of the synthetic data test results not only proved the effectiveness of the proposed method, but also showed that the method has strong noise resistance capability and is suitable for predicting gas-bearing reservoirs with low signal-to-noise ratio seismic data; (2) Based on the drilling data and sedimentary facies characteristics in the test area, the actual 3D seismic data test results showed that there are four wells that match the predicted results, and the distribution boundaries of the gas-bearing reservoirs they depict are relatively clear; (3) In low exploration areas (areas with few or no wells), this method is important as a guide to hydrocarbon exploration.

Mechanical behavior and performance evolution of railway ballast track under dynamic stabiliser based on the hybrid MBD-DEM simulation

The uncertainty in local structure and time-variable compaction of ballast aggregates poses challenges in selecting suitable stabilization methods and parameters for ballast tracks in practical engineering, often resulting in suboptimal operational outcomes. To address this issue comprehensively, the spatial hybrid model of the double dynamic stabilisers-ballast track (MBD-DEM) is proposed in this paper. Subsequent numerical simulations of stabilisation operations are conducted to elucidate ballast vibration, migration behavior and the evolution of ballast track resistance under varying initial conditions of the ballast bed. Initially, leveraging multibody dynamics theory (MBD) and discrete element method (DEM), a MBD model of double stabilisers and a DEM ballast track model containing 6 sleepers were established respectively. Using the coupling interface, the real-time hybrid calculation of these two models is achieved through 6 sleepers in the middle of the ballast bed. Subsequently, the model's accuracy was validated by comparing test data from stabilisation vehicle operations on a Kunming Railway Bureau test line with simulation data from the dynamic double stabilisers-ballast track spatial hybrid model. Finally, utilizing the proposed hybrid model, the dynamic behavior, contact states, and movement rules of ballast particles during stabilisation operations on both loose and initially stable ballast beds (LBB and ISBB) were investigated at the mesoscopic scale. Concurrently, an in-depth analysis of lateral resistance changes in the ballast bed before and after the stabilisation operation is presented. The relevant findings effectively unveil the evolution of dynamic behavior in the ballast bed during stabilisation operations, concluding with practical suggestions for on-site operations. Simulation results illustrate that the rotation of ballast particles in the XY plane beneath each sleeper is predominantly characterized by harmonic rotation. With an increase in the depth of the ballast bed, the range of the rotation axis for combined angular velocity diminishes, leading to the phenomenon of rotation axis deflection. To ensure operational safety and efficiency, a minimum of two-times stabilization operations is recommended for loose ballast beds (LBB), with three-times operations considered optimal. For initially stable ballast beds (ISBB), one-time stabilization operation is basically sufficient.