Unseen Test Research Articles

XGBoost and genetic programming methods for analysing the vitreous transition of the epoxy adhesive

Fibre-reinforced polymer (FRP) composites are increasingly favoured for strengthening existing structures due to their numerous structural benefits. Nevertheless, the performance of such technology is strongly affected by the behaviour of the epoxy resin adhesive layer, which is largely dependent on its curing conditions. This study introduces a deep learning (DL) framework that leverages eXtreme Gradient Boosting (XGBoost) and genetic programming (GP) to comprehensively study the influence of curing scenarios on the vitreous transition of the adhesive. An experimental dataset comprising 160 data points was used to develop predictive models. The XGBoost models exhibited high predictive accuracy for both the onset vitreous transition temperature and the peak tan δ vitreous transition temperature, achieving R 2 values of 0.982 and 0.993 for the unseen test set, respectively. While the GP models exhibited lower predictive accuracy with R 2 values of 0.834 and 0.842, they provided explicit equations that enhance the interpretability of the DL model and facilitate practical application. To make these insights accessible to engineers without programming expertise, a web-based graphical user interface software was developed, incorporating all DL models. Additionally, a feature influence assessment was conducted, providing visual representations of the impact of each feature on the output results, thus enhancing interpretability for engineering applications.

Machine learning identified novel players in lipid metabolism, endosomal trafficking, and iron metabolism of the ALS spinal cord

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease affecting motor neurons. Although genes causing familial cases have been identified, those of sporadic ALS, which occupies the majority of patients, are still elusive. In this study, we adopted machine learning to build binary classifiers based on the New York Genome Center (NYGC) ALS Consortium’s RNA-seq data of the postmortem spinal cord of ALS and non-neurological disease control. The accuracy of the classifiers was greater than 83% and 77% for the training set and the unseen test set, respectively. The classifiers contained 114 genes. Among them, 41 genes have been reported in previous ALS studies, and others are novel in this field. These genes are involved in mitochondrial respiration, lipid metabolism, endosomal trafficking, and iron metabolism, which may promote the progression of ALS pathology.

Predicting microplastic quantities in Indonesian provincial rivers using machine learning models.

Microplastic pollution has surfaced as a critical environmental concern, affecting ecosystems and human health globally. This study explored the application of several machine learning models, including the Tree algorithm, k-Nearest Neighbors (kNN), Random Forest (RF), Linear Regression (LR), Support Vector Machine (SVM), and Neural Networks (NN), to predict microplastic concentrations in the rivers of Indonesia's 24 provinces. By utilizing both environmental and anthropogenic data, the Tree algorithm exhibited the best performance, achieving a coefficient of determination (R2) of 0.838 and a mean absolute percentage error (MAPE) of 0.242 on unseen testing data, thereby highlighting strong predictive capability. Key variables influencing microplastic abundance included annual average temperature, gross domestic product (GDP) per capita and population density. The results underscored the necessity of utilizing comprehensive datasets for effective modeling and highlighted the potential of machine learning to enhance environmental monitoring efforts. This research provides critical insights for policymakers and stakeholders aiming to address the growing issue of microplastic pollution in freshwater systems, providing a foundation for the development of more effective environmental management strategies.

Polymer Biodegradation in Aquatic Environments: A Machine Learning Model Informed by Meta-Analysis of Structure-Biodegradation Relationships.

Polymers are widely produced and contribute significantly to environmental pollution due to their low recycling rates and persistence in natural environments. Biodegradable polymers, while promising for reducing environmental impact, account for less than 2% of total polymer production. To expand the availability of biodegradable polymers, research has explored structure-biodegradability relationships, yet most studies focus on specific polymers, necessitating further exploration across diverse polymers. This study addresses this gap by curating an extensive aerobic biodegradation data set of 74 polymers and 1779 data points drawn from both published literature and 28 sets of original experiments. We then conducted a meta-analysis to evaluate the effects of experimental conditions, polymer structure, and the combined impact of polymer structure and properties on biodegradation. Next, we developed a machine learning model to predict polymer biodegradation in aquatic environments. The model achieved an Rtest2 score of 0.66 using Morgan fingerprints, detailed experimental conditions, and thermal decomposition temperature (Td) as the input descriptors. The model's robustness was supported by a feature importance analysis, revealing that substructure R-O-R in polyethers and polysaccharides positively influenced biodegradation, while molecular weight, Td, substructure -OC(═O)- in polyesters and polyalkylene carbonates, side chains, and aromatic rings negatively impacted it. Additionally, validation against the meta-analysis findings confirmed that predictions for unseen test sets aligned with established empirical biodegradation knowledge. This study not only expands our understanding across diverse polymers but also offers a valuable tool for designing environmentally friendly polymers.

Machine learning-based estimation of crude oil-nitrogen interfacial tension

Accurate estimation of interfacial tension (IFT) between nitrogen and crude oil during nitrogen-based gas injection into oil reservoirs is imperative. The previous research works dealing with prediction of IFT of oil and nitrogen systems consider synthetic oil samples such n-alkanes. In this work, we aim to utilize eight machine learning methods of Decision Tree (DT), AdaBoost (AB), Random Forest (RF), K-nearest Neighbors (KNN), Ensemble Learning (EL), Support Vector Machine (SVM), Convolutional Neural Network (CNN) and Multilayer Perceptron Artificial Neural Network (MLP-ANN) to construct data-driven intelligent models to predict crude oil – nitrogen IFT based upon experimental data of real crude oils samples encountered in underground oil reservoirs. Several statistical indices and graphical approaches are used as accuracy performance indicators. The results show that virtually all the gathered datapoints are suitable for the purpose of model development. The sensitivity analysis indicated that pressure, temperature and crude oil API all negatively affect the IFT, with pressure being the most effective factor. The evaluation study proved that Random Forest is the most accurate developed intelligent model as it was characterized with acceptable R-squared (0.959), mean square error (1.65), average absolute relative error (6.85%) of unseen test datapoints as well as with correct trend prediction of IFT with regard to all input parameters of pressure, temperature and crude oil API. The developed model can be considered an accurate an easy-to-use tool for the prediction of crude oil/N2 IFT values for enhance oil recovery study optimization and upstream reservoir investigations.

Epilepsy Diagnosis from EEG Signals Using Continuous Wavelet Transform-Based Depthwise Convolutional Neural Network Model.

Background/Objectives: Epilepsy is a prevalent neurological disorder characterized by seizures that significantly impact individuals and their social environments. Given the unpredictable nature of epileptic seizures, developing automated epilepsy diagnosis systems is increasingly important. Epilepsy diagnosis traditionally relies on analyzing EEG signals, with recent deep learning methods gaining prominence due to their ability to bypass manual feature extraction. Methods: This study proposes a continuous wavelet transform-based depthwise convolutional neural network (DCNN) for epilepsy diagnosis. The 35-channel EEG signals were transformed into 35-channel images using continuous wavelet transform. These images were then concatenated horizontally and vertically into a single image (seven rows by five columns) using Python's PIL library, which served as input for training the DCNN model. Results: The proposed model achieved impressive performance metrics on unseen test data: 95.99% accuracy, 94.27% sensitivity, 97.29% specificity, and 96.34% precision. Comparative analyses with previous studies and state-of-the-art models demonstrated the superior performance of the DCNN model and image concatenation technique. Conclusions: Unlike earlier works, this approach did not employ additional classifiers or feature selection algorithms. The developed model and image concatenation method offer a novel methodology for epilepsy diagnosis that can be extended to different datasets, potentially providing a valuable tool to support neurologists globally.

Combining simulation and reinforcement learning to reduce food waste in food retail

Extraordinary amounts of fresh produce are never purchased and are discarded as waste. Reinforcement learning (RL) could serve as a means to improve business profits while reducing food waste via control of store pricing and ordering decisions. We present a discrete-event-based simulation framework for food retail which simulates wholesaler, store, and customer interactions. This simulator is critical for driving development and testing of future RL methods. It provides an efficient learning feedback system across a wide gamut of possible scenarios, which cannot be replicated from live observations or pure historical data alone. This is crucial as RL agents cannot learn robust decision-making policies without exposure to many unique scenarios. We evaluate our simulator on a demonstrative case generated from historical consumption and price data using a provided methodology for synthesizing daily demand from monthly and yearly stats. In this demonstrative case, we investigate proximal policy optimization, soft actor–critic, and deep Q networks trained with different reward formulations to decrease food waste and improve profits. These RL methods reduced food waste by 78%–92% on average on an unseen 3-year test period as compared to a baseline mimicking typical food retail waste. Compared to a second popular baseline in literature, the best performing RL algorithm was able to improve profits by up to 12.3%.

Histopathological domain adaptation with generative adversarial networks: Bridging the domain gap between thyroid cancer histopathology datasets.

Deep learning techniques are increasingly being used to classify medical imaging data with high accuracy. Despite this, due to often limited training data, these models can lack sufficient generalizability to predict unseen test data, produced in different domains, with comparable performance. This study focuses on thyroid histopathology image classification and investigates whether a Generative Adversarial Network [GAN], trained with just 156 patient samples, can produce high quality synthetic images to sufficiently augment training data and improve overall model generalizability. Utilizing a StyleGAN2 approach, the generative network produced images with an Fréchet Inception Distance (FID) score of 5.05, matching state-of-the-art GAN results in non-medical domains with comparable dataset sizes. Augmenting the training data with these GAN-generated images increased model generalizability when tested on external data sourced from three separate domains, improving overall precision and AUC by 7.45% and 7.20% respectively compared with a baseline model. Most importantly, this performance improvement was observed on minority class images, tumour subtypes which are known to suffer from high levels of inter-observer variability when classified by trained pathologists.

Optimizing Diabetes Prediction: An Evaluation of Machine Learning Models Through Strategic Feature Selection

Diabetes, a widespread chronic ailment in the United States, imposes significant economic and health burdens, impacting quality of life and life expectancy. This study analyzes a clinical dataset of 253,680 patients from the Behavioral Risk Factor Surveillance System (BRFSS). The dataset encompasses 21 predictors, including high blood pressure, cholesterol, body mass index (BMI), smoking, stroke, heart disease, physical activity, fruit consumption, vegetable consumption, alcohol consumption, insurance coverage, lack of medical visits due to financial constraints, general health, days with mental health issues, days with physical injuries in the past 30 days, difficulties in walking, gender, age, income, and education level. The objective is to balance the training dataset, compare different supervised machine learning models, and identify critical clinical features contributing to diabetes using unsupervised feature selection methods. A total of 34 machine learning models in MATLAB2023a were trained and compared. Quadratic Support Vector Machine (SVM), Coarse Gaussian SVM, and Narrow Neural Networks achieved the highest training accuracy (76.3%), while the Bilayered Neural Network attained 74.7% on an unseen test dataset. Among all, Quadratic SVM demonstrated the best overall performance based on average accuracy, precision, recall, and F1 score. Feature selection highlighted nine key predictors: high blood pressure, high cholesterol, BMI, heart disease, physical activity, general health, recent bodily injuries, mobility issues, and age. A model trained on these features achieved a commendable accuracy of 75.4%, demonstrating the feasibility of a simplified, efficient diagnostic tool with a diagnostic efficacy of 0.7. This study underscores the potential of streamlined models to predict diabetes with fewer parameters while maintaining high accuracy, offering a valuable tool for healthcare diagnostics.

XGBoost-SHAP-based interpretable diagnostic framework for knee osteoarthritis: a population-based retrospective cohort study

ObjectiveTo use routine demographic and clinical data to develop an interpretable individual-level machine learning (ML) model to diagnose knee osteoarthritis (KOA) and to identify highly ranked features.MethodsIn this retrospective, population-based cohort study, anonymized questionnaire data was retrieved from the Wu Chuan KOA Study, Inner Mongolia, China. After feature selections, participants were divided in a 7:3 ratio into training and test sets. Class balancing was applied to the training set for data augmentation. Four ML classifiers were compared by cross-validation within the training set and their performance was further analyzed with an unseen test set. Classifications were evaluated using sensitivity, specificity, positive predictive value, negative predictive value, accuracy, area under the curve(AUC), G-means, and F1 scores. The best model was explained using Shapley values to extract highly ranked features.ResultsA total of 1188 participants were investigated in this study, among whom 26.3% were diagnosed with KOA. Comparatively, XGBoost with Boruta exhibited the highest classification performance among the four models, with an AUC of 0.758, G-means of 0.800, and F1 scores of 0.703. The SHAP method reveals the top 17 features of KOA according to the importance ranking, and the average of the experience of joint pain was recognized as the most important features.ConclusionsOur study highlights the usefulness of machine learning in unveiling important factors that influence the diagnosis of KOA to guide new prevention strategies. Further work is needed to validate this approach.

Prediction of gene expression-based breast cancer proliferation scores from histopathology whole slide images using deep learning

BackgroundIn breast cancer, several gene expression assays have been developed to provide a more personalised treatment. This study focuses on the prediction of two molecular proliferation signatures: an 11-gene proliferation score and the MKI67 proliferation marker gene. The aim was to assess whether these could be predicted from digital whole slide images (WSIs) using deep learning models.MethodsWSIs and RNA-sequencing data from 819 invasive breast cancer patients were included for training, and models were evaluated on an internal test set of 172 cases as well as on 997 cases from a fully independent external test set. Two deep Convolutional Neural Network (CNN) models were optimised using WSIs and gene expression readouts from RNA-sequencing data of either the proliferation signature or the proliferation marker, and assessed using Spearman correlation (r). Prognostic performance was assessed through Cox proportional hazard modelling, estimating hazard ratios (HR).ResultsOptimised CNNs successfully predicted the proliferation score and proliferation marker on the unseen internal test set (ρ = 0.691(p < 0.001) with R2 = 0.438, and ρ = 0.564 (p < 0.001) with R2 = 0.251 respectively) and on the external test set (ρ = 0.502 (p < 0.001) with R2 = 0.319, and ρ = 0.403 (p < 0.001) with R2 = 0.222 respectively). Patients with a high proliferation score or marker were significantly associated with a higher risk of recurrence or death in the external test set (HR = 1.65 (95% CI: 1.05–2.61) and HR = 1.84 (95% CI: 1.17–2.89), respectively).ConclusionsThe results from this study suggest that gene expression levels of proliferation scores can be predicted directly from breast cancer morphology in WSIs using CNNs and that the predictions provide prognostic information that could be used in research as well as in the clinical setting.

Evaluating CNN Architectures for the Automated Detection and Grading of Modic Changes in MRI: A Comparative Study.

Modic changes (MCs) classification system is the most widely used method in magnetic resonance imaging (MRI) for characterizing subchondral vertebral marrow changes. However, it shows a high degree of sensitivity to variations in MRI because of its semiquantitative nature. In 2021, the authors of this classification system further proposed a quantitative and reliable MC grading method. However, automated tools to grade MCs are lacking. This study developed and investigated the performance of convolutional neural network (CNN) in detecting and grading MCs based on their maximum vertical extent. In order to verify performance, we tested CNNs' generalization performance, the performance of CNN with that of junior doctors, and the consistency of junior doctors after AI assistance. A retrospective analysis of 139 patients' MRIs with MCs was conducted and annotated by a spine surgeon. Of the 139 patients, MRIs from 109 patients were acquired using Philips scanners from June 2020 to June 2021, constituting Dataset 1. The remaining 30 patients had MRIs obtained from both Philips and United Imaging scanners from June 2022 to March 2023, forming Dataset 2. YOLOv8 and YOLOv5 were developed in PyCharm using the Python language and based on the PyTorch deep learning framework, data enhancement and transfer learning were applied to enhance model generalization. The model's performance was compared with precision, recall, F1 score, and mAP50. It also tested generalizability and compared it with the junior doctor's performance on the second data set (Dataset 2). Post hoc, the junior doctor graded Dataset 2 with CNN assistance. In addition, the region of interest was displayed using the class activation mapping heat map. On the unseen test set, the YOLOv8 and YOLOv5 models achieved precision of 81.60% and 61.59%, recall of 80.90% and 67.16%, mAP50 of 84.40% and 68.88%, and F1 of 0.81 and 0.60 respectively. On Dataset 2, YOLOv8 and junior doctor achieved precision of 95.1% and 72.5%, recall of 68.3% and 60.6%. In the AI-assisted experiment, agreement between the junior doctor and the senior spine surgeon significantly improved from Cohen's kappa of 0.368-0.681. YOLOv8 in detecting and grading MCs was significantly superior to that of YOLOv5. The performance of YOLOv8 is superior to that of junior doctors, and it can enhance the capabilities of junior doctors and improve the reliability of diagnoses.

AED-PADA:Improving Generalizability of Adversarial Example Detection via Principal Adversarial Domain Adaptation

Adversarial example detection, which can be conveniently applied in many scenarios, is important in the area of adversarial defense. Unfortunately, existing detection methods suffer from poor generalization performance, because their training process usually relies on the examples generated from a single known adversarial attack and there exists a large discrepancy between the training and unseen testing adversarial examples. To address this issue, we propose a novel method, named Adversarial Example Detection via Principal Adversarial Domain Adaptation (AED-PADA). Specifically, our approach identifies the Principal Adversarial Domains (PADs), i.e., a combination of features of the adversarial examples generated by different attacks, which possesses a large portion of the entire adversarial feature space. Subsequently, we pioneer to exploit Multi-source Unsupervised Domain Adaptation in adversarial example detection, with PADs as the source domains. Experimental results demonstrate the superior generalization ability of our proposed AED-PADA. Note that this superiority is particularly achieved in challenging scenarios characterized by employing the minimal magnitude constraint for the perturbations.

Improving subject transfer in EEG classification with divergence estimation

Objective. Classification models for electroencephalogram (EEG) data show a large decrease in performance when evaluated on unseen test subjects. We improve performance using new regularization techniques during model training.Approach. We propose several graphical models to describe an EEG classification task. From each model, we identify statistical relationships that should hold true in an idealized training scenario (with infinite data and a globally-optimal model) but that may not hold in practice. We design regularization penalties to enforce these relationships in two stages. First, we identify suitable proxy quantities (divergences such as Mutual Information and Wasserstein-1) that can be used to measure statistical independence and dependence relationships. Second, we provide algorithms to efficiently estimate these quantities during training using secondary neural network models.Main results. We conduct extensive computational experiments using a large benchmark EEG dataset, comparing our proposed techniques with a baseline method that uses an adversarial classifier. We first show the performance of each method across a wide range of hyperparameters, demonstrating that each method can be easily tuned to yield significant benefits over an unregularized model. We show that, using ideal hyperparameters for all methods, our first technique gives significantly better performance than the baseline regularization technique. We also show that, across hyperparameters, our second technique gives significantly more stable performance than the baseline. The proposed methods require only a small computational cost at training time that is equivalent to the cost of the baseline.Significance. The high variability in signal distribution between subjects means that typical approaches to EEG signal modeling often require time-intensive calibration for each user, and even re-calibration before every use. By improving the performance of population models in the most stringent case of zero-shot subject transfer, we may help reduce or eliminate the need for model calibration.

EpiTCR-KDA: Knowledge Distillation model on Dihedral Angles for TCR-peptide prediction

Abstract Motivation The prediction of the T-cell receptor (TCR) and antigen bindings is crucial for advancements in immunotherapy. However, most current TCR-peptide interaction predictors struggle to perform well on unseen data. This limitation may stem from the conventional use of TCR and/or peptide sequences as input, which may not adequately capture their structural characteristics. Therefore, incorporating the structural information of TCRs and peptides into the prediction model is necessary to improve its generalizability. Results We developed epiTCR-KDA, a new predictor of TCR-peptide binding that utilizes the dihedral angles between the residues of the peptide and the TCR as a structural descriptor. This structural information was integrated into a knowledge distillation model to enhance its generalizability. epiTCR-KDA demonstrated competitive prediction performance, with an AUC of 1.00 for seen data and AUC of 0.91 for unseen data. On public datasets, epiTCR-KDA consistently outperformed other predictors, maintaining a median AUC of 0.93. Further analysis of epiTCR-KDA revealed that the cosine similarity of the dihedral angle vectors between the unseen testing data and training data is crucial for its stable performance. In conclusion, our epiTCR-KDA model represents a significant step forward in developing a highly effective pipeline for antigen-based immunotherapy. Availability and implementation epiTCR-KDA is available on GitHub (https://github.com/ddiem-ri-4D/epiTCR-KDA). Supplementary information Supplementary data are available at Bioinformatics Advances online.

On-site aerodynamics using stereoscopic PIV and deep optical flow learning

We introduce recurrent all-pairs field transforms for stereoscopic particle image velocimetry (RAFT-StereoPIV). Our approach leverages deep optical flow learning to analyze time-resolved and double-frame particle images from on-site measurements, particularly from the ‘Ring of Fire,’ as well as from wind tunnel measurements for fast aerodynamic analysis. A multi-fidelity dataset comprising both Reynolds-averaged Navier–Stokes (RANS) and direct numerical simulation (DNS) was used to train our model. RAFT-StereoPIV outperforms all PIV state-of-the-art deep learning models on benchmark datasets, with a 68 % error reduction on the validation dataset, Problem Class 2, and a 47 % error reduction on the unseen test dataset, Problem Class 1, demonstrating its robustness and generalizability. In comparison with the most recent works in the field of deep learning for PIV, where the main focus was the methodology development and the application was limited to either 2D flow cases or simple experimental data, we extend deep learning-based PIV for industrial applications and three-component two-dimensional (3C2D) velocity estimation. We believe that this study brings the field of experimental fluid dynamics one step closer to the long-term goal of having experimental measurement systems that can be used for fast flow field estimation.

HPB O06 - A computer tomography-based radiomics signature predicts overall survival following curative resection of colorectal liver metastasis: A multicentre, retrospective study

Abstract Background Colorectal cancer liver metastasis (CRLM) has a heterogeneous outcome and improved prognostic biomarkers are needed to aid patient-tailored management. Here, we assess the value of Computer Tomography (CT)-guided radiomics in the prognostication of post-resection CRLM using data from two large tertiary referral centres in the UK. Method Patient data was extracted from prospectively maintained databases of patients undergoing CRLM resection at two centres. CRLM and regions of interest from background normal liver were segmented. Radiomics features were extracted from portal venous phase CT images using open-source software. Feature selection techniques were applied. Machine learning models incorporating radiomic or clinical features, and their combination, were developed and assessed for mortality prediction at fixed time points. Cox proportional hazard regression was utilised for risk score generation to classify patients as high or low-risk for overall survival estimation. The models were evaluated on unseen testing data. Results Radiomics features (269) were generated from 959 metastases in 399 patients across the two sites. Models with combined clinical and radiomic features performed similarly to those with radiomic features alone, and often outperformed models with clinical features alone demonstrating good discrimination and risk prediction at 2 and 3 years. The computed radiomic scores allowed us to generate Kaplan–Meier curves which showed CRLM patients in the high-risk group had a lower overall survival vs the low-risk group(p&amp;lt;0.05). This risk-score independently predicted mortality in our multivariate Cox proportional-hazards model. Conclusion We demonstrate a robust combination of CT-based radiomics features can predict CRLM outcomes across two different hospital sites. Our radiomic risk-scores predicts overall survival and can prognosticate patients into high and low risk groups. These findings support the prognostic utility of CT-based radiomics in patients undergoing resection of CRLM and provide rationale for further investigation in a prospective, multi-centre setting.

Quantum machine learning enhanced laser speckle analysis for precise speed prediction

Laser speckle contrast imaging (LSCI) is an optical technique used to assess blood flow perfusion by modeling changes in speckle intensity, but it is generally limited to qualitative analysis due to difficulties in absolute quantification. Three-dimensional convolutional neural networks (3D CNNs) enhance the quantitative performance of LSCI by excelling at extracting spatiotemporal features from speckle data. However, excessive downsampling techniques can lead to significant information loss. To address this, we propose a hybrid quantum–classical 3D CNN framework that leverages variational quantum algorithms (VQAs) to enhance the performance of classical models. The proposed framework employs variational quantum circuits (VQCs) to replace the 3D global pooling layer, enabling the model to utilize the complete 3D information extracted by the convolutional layers for feature integration, thereby enhancing velocity prediction performance. We perform cross-validation on experimental LSCI speckle data and demonstrate the superiority of the hybrid models over their classical counterparts in terms of prediction accuracy and learning stability. Furthermore, we evaluate the models on an unseen test set and observe that the hybrid models outperform the classical models with up to 14.8% improvement in mean squared error (MSE) and up to 26.1% improvement in mean absolute percentage error (MAPE) evaluation metrics. Finally, our qualitative analysis shows that the hybrid models offer substantial improvements over classical models in predicting blood flow at both low and high velocities. These results indicate that the hybrid models possess more powerful learning and generalization capabilities.

ANN prediction of mode I fracture energy in epoxy adhesive joints: Adherend, adhesive, and strain rate effects

In this investigation, the mode I fracture behavior of an epoxy adhesive joint has been investigated to probe the combined effects of adherend Young's modulus and thickness, strain rate, and adhesive length. Experimental fracture testing was conducted on double-cantilever beam (DCB) specimens under three various strain rate thresholds, namely, quasi-static (0.0005 s−1), low (0.015 s−1), and intermediate (0.5 s−1). The specimens were constructed with varying adherend materials (i.e., copper, aluminum), thickness (12–20 mm), and adhesive lengths (25–75 mm). The fracture load was measured in each experiment, and the critical strain energy release rate, JCi, was calculated via a finite element analysis (FEA) in each case. Results with a 95 % confidence interval showed adherend thickness had a negligible effect within the investigated range (p-value = 0.462). JCi was significantly impacted by adhesive length (p-value = 0.009) and two other investigated parameters (p-values = 0.000). The development of artificial neural networks (ANNs) with Levenberg-Marquardt (LM) led to the prediction of JCi based on the examined parameters. With a mean absolute percentage error (MAPE) of 8.9 % and a mean squared error (MSE) of 683 J2/m4 on unseen test data, the ANN model built in this study was able to predict the JCi in epoxy adhesive joints, indicating its potential to produce an accurate prediction for JCi in epoxy adhesive joints. This work offers insights for precise fracture behavior prediction under mode I stress conditions as well as some helpful information that can be utilized to optimize adhesive joint design.

Abstract 4116410: A Novel Deep Learning Approach for Prediction of Right Heart Failure After Left Ventricular Assist Device Implantation using Pulmonary Artery Pressure Tracings

Introduction: Left ventricular assist device (LVAD) offers a lifeline for advanced heart failure patients, but ~30% experience post-operative right heart failure (RHF) with significant morbidity and mortality. Current methods for predicting RHF risk have limited accuracy. This study is the first of its kind to explore the feasibility of using a convolutional neural network (CNN) deep learning model with pulmonary artery pressure (PAP) tracings acquired during pre-operative hemodynamic assessment to identify patients at risk for early RHF. Methods: A CNN model was pre-trained on PAP tracings from the MIMIC-III Waveform Database to leverage transferable knowledge and improve model performance. The model learned general features and temporal relationships in hemodynamic tracing morphology for RHF prediction. It was then adapted for classification by modifying the output layers. Post-LVAD RHF was definitively adjudicated for each instance included from our 246 patient single center LVAD database. RHF was adjudicated using the 2020 Academic Research Consortium defintion. The database included 205 non-RHF instances and 41 RHF instances. We developed a novel signal processing method utilizing computer vision to extract PAP waveform data from pre-operative right heart catheterization reports. Data was reviewed to ensure accurate pressure tracing representation. SMOTE-ENN (synthetic oversampling) and class weighting addressed dataset imbalance during training to enhance model generalizability and RHF prediction accuracy. Results: The model achieved strong performance in classifying RHF from non-RHF outcomes, as evidenced by consistent AUC scores of 0.87 and 0.85 on validation and unseen test data, respectively. These findings indicate the potential for hemodynamic tracings to predict RHF after LVAD implantation. Notably, the model maintained its performance on unseen data, highlighting its generalizability. Conclusions: This study explored the feasibility of using a deep learning model for RHF risk stratification in patients with LVADs. The model, pre-trained on PAP tracings, achieved promising classification performance assessed by AUC. Further validation with larger, more diverse datasets will refine model performance and enhance generalizability for clinical application. Leveraging hemodynamic tracings within a deep learning framework holds promise for improved clinical outcome prediction.