Searcharxiv⌕ Search

arXiv subjects

Emma Lejeune

Publications and source records attributed to Emma Lejeune.

At least 19 recordsLinked to original sources
arXiv2026

Structured AI Demonstrations and Student LLM Use in Engineering Mechanics: Study Design and Preliminary Results

The rapid integration of large language models (LLMs) into undergraduate education presents an urgent challenge for engineering instructors. Despite widespread student adoption, there remains a critical lack of domain-specific empirical evidence to guide pedagogical policies and classroom interventions. This manuscript presents a descriptive study design and preliminary findings from an undergraduate engineering mechanics course conducted in Spring 2026. We detail a reproducible survey instrument used to capture student AI usage patterns, attitudes, and verification practices, which are subsequently linked to academic performance metrics. Additionally, we document a deployable sequence of nine structured, instructor-led AI demonstrations designed to model strategic LLM delegation and evaluation. While our preliminary data highlight shifting student behaviors and complex relationships between AI reliance and course outcomes, the primary contribution of this work is the provision of an open-access methodological framework. By making our complete study design, survey tools, and demonstration materials publicly available, we urge other engineering educators to collect and share similar empirical data. Navigating this unprecedented technological shift will require a collaborative, evidence-based approach to fully understand its long-term impacts on student learning.

cs.CY↗
arXiv2026

Quantifying the Spatiotemporal Dynamics of Engineered Cardiac Microbundles

Brightfield time-lapse imaging is widely used in cardiac tissue engineering, yet the absence of standardized, interpretable analytical frameworks limits reproducibility and cross-platform comparison. We present an open, scalable computational pipeline for quantifying spatiotemporal contractile dynamics in microscopy videos of human induced pluripotent stem cell-derived cardiac microbundles. Building on our open-source tools "MicroBundleCompute" and "MicroBundlePillarTrack," we define a suite of 16 interpretable structural, functional, and spatiotemporal metrics that capture tissue deformation, synchrony, and heterogeneity. The framework integrates full-field displacement tracking, strain reconstruction, spatial registration, dimensionality reduction, and topology-based vector-field analysis within a unified workflow. Applied to a dataset of 670 cardiac microbundles spanning 20 experimental conditions, the pipeline reveals continuous variation in contractile phenotypes rather than discrete condition-specific clustering, with intra-condition variability often exceeding inter-condition differences. Redundancy analysis identifies a reduced core set of 10 metrics that retain most informational content while minimizing multicollinearity. Analysis of denoised displacement fields shows that contraction is dominated by a global isotropic mode, with localized saddle-type deformation patterns present in approximately half of the samples. All software and workflows are released openly to enable reproducible, scalable analysis of dynamic tissue mechanics.

q-bio.QM↗
arXiv2026

Interdisciplinary Workshop on Mechanical Intelligence: Summary Report

This report provides a summary of the outcomes of the Interdisciplinary Workshop on Mechanical Intelligence held in 2024. Mechanical Intelligence (MI) represents the phenomenon that novel structural features of material/biological/robotic systems can encode intelligence through responsiveness, adaptivity, memory, and learning in the mechanical structure itself. This is in contrast to computational intelligence, wherein the intelligence functions occur through electrical signaling and computer code. The two-day workshop was held at NSF headquarters on May 30-31 and included 38 invited academic researcher participants, and 8 program officers from the NSF. The workshop was structured around active small and large group discussions in groups of 4-5 and 9-10 with the goal of addressing topical questions on MI. Working groups entered notes into shared presentation slides for each discussion session and presented their outcomes in a final presentation on the last day. Here we summarize the overall outcomes of the workshop.

cs.RO↗
arXiv2026

Information Propagation and Encoding in Solids: A Quantitative Approach Towards Mechanical Intelligence

Engineered systems typically separate mechanical function from information processing, whereas biological systems can exploit physical structure as a medium for information processing and computation. Motivated by this contrast, recent work in mechanics has explored embedding information-processing capabilities directly into mechanical structures. However, quantitative frameworks for evaluating such capabilities remain limited. Here we address a foundational question: how does information propagate through a solid body? Using elastic bodies as a model system, we apply information-theoretic tools to treat an elastic domain as an information encoder and quantify how information transmits from applied loads to discrete sensor locations. We further connect these measures to familiar mechanical phenomena, including Saint-Venant's effect and principal stress lines. Moving toward design, we show how geometry and architected materials can tune transmission, enabling elastic domains to either transmit or block information. Overall, this work advances quantifiable metrics and benchmark tasks for mechanical intelligence, supporting comparable designs of mechanically embodied information processing.

cs.IT↗
arXiv2025

FEM-Bench: A Structured Scientific Reasoning Benchmark for Evaluating Code-Generating LLMs

As LLMs advance their reasoning capabilities about the physical world, the absence of rigorous benchmarks for evaluating their ability to generate scientifically valid physical models has become a critical gap. Computational mechanics, which develops and applies mathematical models and numerical methods to predict the behavior of physical systems under forces, deformation, and constraints, provides an ideal foundation for structured scientific reasoning evaluation. Problems follow clear mathematical structure, enforce strict physical and numerical constraints, and support objective verification. The discipline requires constructing explicit models of physical systems and reasoning about geometry, spatial relationships, and material behavior, connecting directly to emerging AI goals in physical reasoning and world modeling. We introduce FEM-Bench, a computational mechanics benchmark designed to evaluate the ability of LLMs to generate correct finite element method (FEM) and related code. FEM-Bench 2025 contains a suite of introductory but nontrivial tasks aligned with material from a first graduate course on computational mechanics. These tasks capture essential numerical and physical modeling challenges while representing only a small fraction of the complexity present in the discipline. Despite their simplicity, state-of-the-art LLMs do not reliably solve all of them. In a five attempt run, the best performing model at function writing, Gemini 3 Pro, completed 30/33 tasks at least once and 26/33 tasks all five times. The best performing model at unit test writing, GPT-5, had an Average Joint Success Rate of 73.8%. Other popular models showed broad performance variation. FEM-Bench establishes a structured foundation for evaluating AI-generated scientific code, and future iterations will incorporate increasingly sophisticated tasks to track progress as models evolve.

cs.LG↗
arXiv2025

SarcGraph for High-Throughput Regional Analysis of Sarcomere Organization and Contractile Function in 2D Cardiac Muscle Bundles

Timelapse images of human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) provide rich information on cell structure and contractile function. However, it is challenging to reproducibly generate tissue samples and conduct scalable experiments with these cells. The two-dimensional cardiac muscle bundle (2DMB) platform helps address these limitations by standardizing tissue geometry, resulting in physiologic, uniaxial contractions of discrete tissues on an elastomeric substrate with stiffness similar to the heart. 2DMBs are highly conducive to sarcomere imaging using fluorescent reporters, but, due to their larger and more physiologic sarcomere displacements and velocities, prior sarcomere-tracking pipelines have been unreliable. Here, we present adaptations to SarcGraph, an open-source Python package for sarcomere detection and tracking, that enable automated analysis of high-frame-rate 2DMB recordings. Key modifications to the pipeline include: 1) switching to a frame-by-frame sarcomere detection approach and automating tissue segmentation with spatial partitioning, 2) performing Gaussian Process Regression for signal denoising, and 3) incorporating an automatic contractile phase detection pipeline. These enhancements enable the extraction of structural organization and functional contractility metrics for both the whole 2DMB tissue and distinct tissue regions, both in a fully automated manner. We complement this software release with a dataset of 130 example movies of baseline and drug-treated samples disseminated through the Harvard Dataverse. By providing open-source tools and datasets, we aim to enable high-throughput analysis of engineered cardiac tissues and advance collective progress within the hiPSC-CM research community.

q-bio.QM↗
arXiv2025

Towards Robust Surrogate Models: Benchmarking Machine Learning Approaches to Expediting Phase Field Simulations of Brittle Fracture

Data driven approaches have the potential to make modeling complex, nonlinear physical phenomena significantly more computationally tractable. For example, computational modeling of fracture is a core challenge where machine learning techniques have the potential to provide a much needed speedup that would enable progress in areas such as mutli-scale modeling and uncertainty quantification. Currently, phase field modeling (PFM) of fracture is one such approach that offers a convenient variational formulation to model crack nucleation, branching and propagation. To date, machine learning techniques have shown promise in approximating PFM simulations. However, most studies rely on overly simple benchmarks that do not reflect the true complexity of the fracture processes where PFM excels as a method. To address this gap, we introduce a challenging dataset based on PFM simulations designed to benchmark and advance ML methods for fracture modeling. This dataset includes three energy decomposition methods, two boundary conditions, and 1,000 random initial crack configurations for a total of 6,000 simulations. Each sample contains 100 time steps capturing the temporal evolution of the crack field. Alongside this dataset, we also implement and evaluate Physics Informed Neural Networks (PINN), Fourier Neural Operators (FNO) and UNet models as baselines, and explore the impact of ensembling strategies on prediction accuracy. With this combination of our dataset and baseline models drawn from the literature we aim to provide a standardized and challenging benchmark for evaluating machine learning approaches to solid mechanics. Our results highlight both the promise and limitations of popular current models, and demonstrate the utility of this dataset as a testbed for advancing machine learning in fracture mechanics research.

cs.LG↗
arXiv2025

Quantifying HiPSC-CM Structural Organization at Scale with Deep Learning-Enhanced SarcGraph

In cardiac cells, structural organization is an important indicator of cell maturity and healthy function. Healthy cardiomyocytes exhibit well-aligned morphology with densely packed and organized sarcomeres. Immature or diseased cardiomyocytes typically lack this organized structure. Critically, human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) offer a valuable model for studying human cardiac cells in a controlled environment. However, these cells often exhibit a disorganized structure. In this work, we extend the SarcGraph computational framework -- designed to assess the structural and functional behavior of hiPSC-CMs -- to better accommodate the structural features of immature cells. There are two key enhancements: (1) incorporating a deep learning-based z-disc classifier, and (2) introducing a novel ensemble graph-scoring approach. These modification significantly reduced false positive sarcomere detections in immature cells, and resulted in the detection of longer myofibrils in mature samples. With this enhanced framework, we analyze an open-source dataset published by the Allen Institute for Cell Science, where, for the first time, we are able to extract key structural features from these data using information from each individually detected sarcomere. Not only are we able to use these structural features to predict expert scores, but we are also able to use these structural features to identify bias in expert scoring and offer an alternative unsupervised learning approach based on explainable clustering. These results demonstrate the efficacy of our modified SarcGraph in extracting biologically meaningful features, enabling a deeper understanding of hiPSC-CM structural integrity. By making our code and tools open-source, we aim to empower the broader cardiac research community and foster further development of computational tools for cardiac tissue analysis.

q-bio.QM↗
arXiv2025

Towards understanding structure-function relationships in random fiber networks

Random fiber networks form the structural foundation of numerous biological tissues and engineered materials. From a mechanics perspective, understanding the structure-function relationships of random fiber networks is particularly interesting because when external force is applied to these networks, only a small subset of fibers will actually carry the majority of the load. Specifically, these load-bearing fibers propagate through the network to form load paths, also called force chains. However, the relationship between fiber network geometric structure, force chains, and the overall mechanical behavior of random fiber network structures remains poorly understood. To this end, we implement a finite element model of random fiber networks with geometrically exact beam elements, and use this model to explore random fiber network mechanical behavior. Our focus is twofold. First, we explore the mechanical behavior of single fiber chains and random fiber networks. Second, we propose and validate an interpretable analytical approach to predicting fiber network mechanics from structural information alone. Key findings include insight into the critical strain-stiffening transition point for single fiber chains and fiber networks generated from a Voronoi diagram, and a connection between force chains and the distance-weighted graph shortest paths that arise by treating fiber networks as spatial graph structures. This work marks an important step towards mapping the structure-function relationships of random fiber networks undergoing large deformations. Additionally, with our code distributed under open-source licenses, we hope that future researchers can directly build on our work to address related problems beyond the scope defined here.

cond-mat.soft↗
arXiv2024

MicroBundlePillarTrack: A Python package for automated segmentation, tracking, and analysis of pillar deflection in cardiac microbundles

Movies of human induced pluripotent stem cell (hiPSC)-derived engineered cardiac tissue (microbundles) contain abundant information about structural and functional maturity. However, extracting these data in a reproducible and high-throughput manner remains a major challenge. Furthermore, it is not straightforward to make direct quantitative comparisons across the multiple in vitro experimental platforms employed to fabricate these tissues. Here, we present "MicroBundlePillarTrack," an open-source optical flow-based package developed in Python to track the deflection of pillars in cardiac microbundles grown on experimental platforms with two different pillar designs ("Type 1" and "Type 2" design). Our software is able to automatically segment the pillars, track their displacements, and output time-dependent metrics for contractility analysis, including beating amplitude and rate, contractile force, and tissue stress. Because this software is fully automated, it will allow for both faster and more reproducible analyses of larger datasets and it will enable more reliable cross-platform comparisons as compared to existing approaches that require manual steps and are tailored to a specific experimental platform. To complement this open-source software, we share a dataset of 1,540 brightfield example movies on which we have tested our software. Through sharing this data and software, our goal is to directly enable quantitative comparisons across labs, and facilitate future collective progress via the biomedical engineering open-source data and software ecosystem.

q-bio.QM↗
arXiv2024

Machine Learning-Guided Design of Non-Reciprocal and Asymmetric Elastic Chiral Metamaterials

There has been significant recent interest in the mechanics community to design structures that can either violate reciprocity, or exhibit elastic asymmetry or odd elasticity. While these properties are highly desirable to enable mechanical metamaterials to exhibit novel wave propagation phenomena, it remains an open question as to how to design passive structures that exhibit both significant non-reciprocity and elastic asymmetry. In this paper, we first define several design spaces for chiral metamaterials leveraging specific design parameters, including the ligament contact angles, the ligament shape, and circle radius. Having defined the design spaces, we then leverage machine learning approaches, and specifically Bayesian optimization, to determine optimally performing designs within each design space satisfying maximal non-reciprocity or stiffness asymmetry. Finally, we perform multi-objective optimization by determining the Pareto optimum and find chiral metamaterials that simultaneously exhibit high non-reciprocity and stiffness asymmetry. Our analysis of the underlying mechanisms reveals that chiral metamaterials that can display multiple different contact states under loading in different directions are able to simultaneously exhibit both high non-reciprocity and stiffness asymmetry. Overall, this work demonstrates the effectiveness of employing ML to bring insights to a novel domain with limited prior information, and more generally will pave the way for metamaterials with unique properties and functionality in directing and guiding mechanical wave energy.

physics.app-ph↗
arXiv2024

MicroBundleCompute: Automated segmentation, tracking, and analysis of subdomain deformation in cardiac microbundles

Advancing human induced pluripotent stem cell derived cardiomyocyte (hiPSC-CM) technology will lead to significant progress ranging from disease modeling, to drug discovery, to regenerative tissue engineering. Yet, alongside these potential opportunities comes a critical challenge: attaining mature hiPSC-CM tissues. At present, there are multiple techniques to promote maturity of hiPSC-CMs including physical platforms and cell culture protocols. However, when it comes to making quantitative comparisons of functional behavior, there are limited options for reliably and reproducibly computing functional metrics that are suitable for direct cross-system comparison. In addition, the current standard functional metrics obtained from time-lapse images of cardiac microbundle contraction reported in the field (i.e., post forces, average tissue stress) do not take full advantage of the available information present in these data (i.e., full-field tissue displacements and strains). Thus, we present "MicroBundleCompute," a computational framework for automatic quantification of morphology-based mechanical metrics from movies of cardiac microbundles. Briefly, this computational framework offers tools for automatic tissue segmentation, tracking, and analysis of brightfield and phase contrast movies of beating cardiac microbundles. It is straightforward to implement, requires little to no parameter tuning, and runs quickly on a personal computer. In this paper, we describe the methods underlying this computational framework, show the results of our extensive validation studies, and demonstrate the utility of exploring heterogeneous tissue deformations and strains as functional metrics. With this manuscript, we disseminate "MicroBundleCompute" as an open-source computational tool with the aim of making automated quantitative analysis of beating cardiac microbundles more accessible to the community.

q-bio.QM↗
arXiv2023

Segmenting mechanically heterogeneous domains via unsupervised learning

From biological organs to soft robotics, highly deformable materials are essential components of natural and engineered systems. These highly deformable materials can have heterogeneous material properties, and can experience heterogeneous deformations with or without underlying material heterogeneity. Many recent works have established that computational modeling approaches are well suited for understanding and predicting the consequences of material heterogeneity and for interpreting observed heterogeneous strain fields. In particular, there has been significant work towards developing inverse analysis approaches that can convert observed kinematic quantities (e.g., displacement, strain) to material properties and mechanical state. Despite the success of these approaches, they are not necessarily generalizable and often rely on tight control and knowledge of boundary conditions. Here, we will build on the recent advances (and ubiquity) of machine learning approaches to explore alternative approaches to detect patterns in heterogeneous material properties and mechanical behavior. Specifically, we will explore unsupervised learning approaches to clustering and ensemble clutering to identify heterogeneous regions. Overall, we find that these approaches are effective, yet limited in their abilities. Through this initial exploration (where all data and code is published alongside this manuscript), we set the stage for future studies that more specifically adapt these methods to mechanical data.

cs.LG↗
arXiv2023

A Data-Driven Computational Model for Engineered Cardiac Microtissues

Engineered heart tissues (EHTs) present a potential solution to some of the current challenges in the treatment of heart disease; however, the development of mature, adult-like cardiac tissues remains elusive. Mechanical stimuli have been observed to improve whole-tissue function and cardiomyocyte (CM) maturation, although our ability to fully utilize these mechanisms is hampered, in part, by our incomplete understanding of the mechanobiology of EHTs. In this work, we leverage the experimental data produced by a mechanically tunable experimental setup to generate tissue-specific computational models of EHTs. Using imaging and functional data, our modeling pipeline generates models with tissue-specific ECM and myofibril structure, allowing us to estimate CM active stress. We use this experimental and modeling pipeline to study different mechanical environments, where we contrast the force output of the tissue with the computed active stress of CMs. We show that the significant differences in measured experimental forces can largely be explained by the levels of myofibril formation achieved by the CMs in the distinct mechanical environments, with active stress showing more muted variations across conditions. The presented model also enables us to dissect the relative contributions of myofibrils and extracellular matrix to tissue force output, a task difficult to address experimentally. These results highlight the importance of tissue-specific modeling to augment EHT experiments, providing deeper insights into the mechanobiology driving EHT function.

q-bio.TO↗
arXiv2023

Locality sensitive hashing via mechanical behavior

From healing wounds to maintaining homeostasis in cyclically loaded tissue, living systems have a phenomenal ability to sense, store, and respond to mechanical stimuli. Broadly speaking, there is significant interest in designing engineered systems to recapitulate this incredible functionality. In engineered systems, we have seen significant recent computationally driven advances in sensing and control. And, there has been a growing interest - inspired in part by the incredible distributed and emergent functionality observed in the natural world - in exploring the ability of engineered systems to perform computation through mechanisms that are fundamentally driven by physical laws. In this work, we focus on a small segment of this broad and evolving field: locality sensitive hashing via mechanical behavior. Specifically, we will address the question: can mechanical information (i.e., loads) be transformed by mechanical systems (i.e., converted into sensor readouts) such that the mechanical system meets the requirements for a locality sensitive hash function? Overall, we not only find that mechanical systems are able to perform this function, but also that different mechanical systems vary widely in their efficacy at this task. Looking forward, we view this work as a starting point for significant future investigation into the design and optimization of mechanical systems for conveying mechanical information for downstream computing.

eess.SP↗
arXiv2023

Investigating Deep Learning Model Calibration for Classification Problems in Mechanics

Recently, there has been a growing interest in applying machine learning methods to problems in engineering mechanics. In particular, there has been significant interest in applying deep learning techniques to predicting the mechanical behavior of heterogeneous materials and structures. Researchers have shown that deep learning methods are able to effectively predict mechanical behavior with low error for systems ranging from engineered composites, to geometrically complex metamaterials, to heterogeneous biological tissue. However, there has been comparatively little attention paid to deep learning model calibration, i.e., the match between predicted probabilities of outcomes and the true probabilities of outcomes. In this work, we perform a comprehensive investigation into ML model calibration across seven open access engineering mechanics datasets that cover three distinct types of mechanical problems. Specifically, we evaluate both model and model calibration error for multiple machine learning methods, and investigate the influence of ensemble averaging and post hoc model calibration via temperature scaling. Overall, we find that ensemble averaging of deep neural networks is both an effective and consistent tool for improving model calibration, while temperature scaling has comparatively limited benefits. Looking forward, we anticipate that this investigation will lay the foundation for future work in developing mechanics specific approaches to deep learning model calibration.

cs.LG↗
arXiv2022

Towards out of distribution generalization for problems in mechanics

There has been a massive increase in research interest towards applying data driven methods to problems in mechanics. While traditional machine learning (ML) methods have enabled many breakthroughs, they rely on the assumption that the training (observed) data and testing (unseen) data are independent and identically distributed (i.i.d). Thus, traditional ML approaches often break down when applied to real world mechanics problems with unknown test environments and data distribution shifts. In contrast, out-of-distribution (OOD) generalization assumes that the test data may shift (i.e., violate the i.i.d. assumption). To date, multiple methods have been proposed to improve the OOD generalization of ML methods. However, because of the lack of benchmark datasets for OOD regression problems, the efficiency of these OOD methods on regression problems, which dominate the mechanics field, remains unknown. To address this, we investigate the performance of OOD generalization methods for regression problems in mechanics. Specifically, we identify three OOD problems: covariate shift, mechanism shift, and sampling bias. For each problem, we create two benchmark examples that extend the Mechanical MNIST dataset collection, and we investigate the performance of popular OOD generalization methods on these mechanics-specific regression problems. Our numerical experiments show that in most cases, while the OOD generalization algorithms perform better compared to traditional ML methods on these OOD problems, there is a compelling need to develop more robust OOD generalization methods that are effective across multiple OOD scenarios. Overall, we expect that this study, as well as the associated open access benchmark datasets, will enable further development of OOD generalization methods for mechanics specific regression problems.

stat.ML↗
arXiv2022

Enhancing Mechanical Metamodels with a Generative Model-Based Augmented Training Dataset

Modeling biological soft tissue is complex in part due to material heterogeneity. Microstructural patterns, which play a major role in defining the mechanical behavior of these tissues, are both challenging to characterize, and difficult to simulate. Recently, machine learning-based methods to predict the mechanical behavior of heterogeneous materials have made it possible to more thoroughly explore the massive input parameter space associated with heterogeneous blocks of material. Specifically, we can train machine learning (ML) models to closely approximate computationally expensive heterogeneous material simulations where the ML model is trained on a dataset of simulations that capture the range of spatial heterogeneity present in the material of interest. However, when it comes to applying these techniques to biological tissue more broadly, there is a major limitation: the relevant microstructural patterns are both challenging to obtain and difficult to analyze. Consequently, the number of useful examples available to characterize the input domain under study is limited. In this work, we investigate the efficacy of ML-based generative models as well as procedural methods as a tool for augmenting limited input pattern datasets. We find that a Style-based Generative Adversarial Network with adaptive discriminator augmentation is able to successfully leverage just 1,000 example patterns to create the most authentic generated patterns. In general, diverse generated patterns with adequate resemblance to the real patterns can be used as inputs to finite element simulations to meaningfully augment the training dataset. To enable this methodological contribution, we have created an open access dataset of Finite Element Analysis simulations based on Cahn-Hilliard patterns. We anticipate that future researchers will be able to leverage this dataset and build on the work presented here.

cs.LG↗