Research Overview

The most pressing challenge in brain mechanics today is understanding the complex relationship between mechanical forces and brain function. The brain, a soft and highly heterogeneous organ, exhibits unique mechanical properties that remain poorly understood. Even slight mechanical alterations, such as those caused by injury, disease, or aging, can lead to significant disruptions in neural activity. To address this challenge, fundamental research is required to investigate these mechanical interactions across multiple scales, from the cellular to the organ level. Similarly, in the field of materials design, the challenge lies in exploring the vast, uncharted space of material combinations and predicting their properties with accuracy. While recent advancements have been made, current approaches are slow and resource-intensive, necessitating new computational tools to expedite material discovery.

It is my firmly held belief that we are on the cusp of transformative breakthroughs in both brain mechanics and AI-driven materials design. In brain mechanics, I am optimistic that by advancing our understanding of how mechanical forces influence brain function, we will develop novel therapies for neurodegenerative diseases and brain injuries. In materials design, I am confident that artificial intelligence will play a pivotal role in accelerating the discovery of innovative materials, which could lead to groundbreaking advancements in sustainable energy, advanced manufacturing, and extreme environment applications.

In my research group, we focus on developing and applying multiscale and multiphysics models to simulate mechanical forces within brain tissue, integrating experimental data with computational predictions. These models allow us to investigate how cellular-level changes impact overall brain function, paving the way for new protective and therapeutic strategies. For materials design, we harness machine learning algorithms to analyze large datasets and predict new materials with tailored properties. By combining multiscale modeling with data-driven approaches, we aim to revolutionize the discovery process, creating materials optimized for real-world applications before they are even synthesized. With these research approaches, I remain optimistic that both fields are on the brink of major discoveries, and the right computational tools will allow us to unlock new solutions to improve human health and technological progress.

Research Thrust I: Data-Driven Brain Mechanics

Convoluted cortical folding, characterized by convex gyri and concave sulci; and neuronal wiring, characterized by short- and long-distance axonal connections; are two prominent attributes of the human brain. Many studies have shown that knowledge of cortical folding is key to interpreting the normal development of the human brain during the early stages of growth. Cognitive or physiological difficulties and problems, e.g., epilepsy, retardation, autism and schizophrenia, are consequences of abnormal cortical folding in the fetal stage fetal stage. However, it is very challenging to answer why the primary cortical convolution organization across subjects within each species is highly correlated and consistent rather than random and what factors count for this consistency as regulators. The intrinsic relationship between these two general cross-species attributes, as well as the underlying principles that sculpt the structural architecture of the cerebral cortex, remains poorly understood. Therefore, there is a critical need to explore the fundamental mechanical principles of normal cortical folding and provide novel diagnostics and treatments of neurological disorders during early brain development.

1. Jixin Hou, Xianyan Chen, Taotao Wu, Ellen Kuhl, Xianqiao Wang, “Automated Data-Driven Discovery of Material Models Based on Symbolic Regression: A Case Study on Human Brain Cortex”, Acta Biomaterialia, in press, 2024.
2. Nicholas Filla, Jixin Hou, Tianming Liu, Silvia Budday, Xianqiao Wang, “Accuracy Meets Simplicity: a Constitutive Model for Heterogenous Brain Tissue”, Journal of the Mechanical Behavior of Biomedical Materials, 106271, 2023.
3. Poorya Chavoshnejad, Liangjun Chen, Jixin Hou, Dajiang Zhu, Tianming Liu, Gang Li, Mir Jalil Razavi, Xianqiao Wang, “An Integrated Finite Element Method and Machine Learning Algorithm for Brain Morphology Prediction”, Cerebral Cortex, 33(15): 9354-9366, 2023.
4. Mir Jalil Razavi, Tianming Liu, Xianqiao Wang. “Mechanism Exploration of 3-Hinge Gyral Formation and Pattern Recognition”, Cerebral Cortex Communications, 2(3): tgab044, 2021.

Collaborators: Dr. Ellen Kuhl (Stanford U); Dr. Tianming Liu (UGA); Dr. Gang Li (UNC); Dr. Jalil Razavi (Binghamton U); Dr. Dajiang Zhu (UTA); Dr. Tuo Zhang (NPU)

Research Thrust II: Machine-Learning Enhanced Mechanics of Biological Systems

Blood clot formation is an essential step in maintaining physiological hemostasis. However, undesired clots or thrombi generated under pathological conditions may partially or totally obstruct blood vessels, causing life-threatening crises, such as myocardial infarction, ischemic stroke, and pulmonary embolism, which collectively are responsible for 1 in 4 deaths worldwide. Biomechanical properties of a thrombus, which are dictated by its micro-structural features and composition, play a crucial role in determining thrombus fates, i.e., occlude, persist, or embolize, in the circulatory system. Extensive experimental studies have been performed to unveil the structural elements of various types of thrombi at a microscopic scale, and to examine the mechanical responses of thrombi under different external loadings at a macroscale, but there is a lack of tools to quantitatively bridge the findings and data from these multiscale experiments to improve our understanding of the links among the micro-structure and composition of a thrombus and its biomechanical properties and mechanical behaviors within the blood flow. In addition, many characteristic features associated with thrombus contraction, a dynamic process to reinforce the function of a thrombus in hemostasis by modulating its structure and biomechanical properties, have not been fully understood because they cannot be observed or measured experimentally. The goal of this project is to build a deep learning-enhanced multiscale computational framework that can blend the micro-structural features, and composition of thrombi with their biomechanical properties for predictive modeling of thrombus fate in circulation. Moreover, the proposed multiscale models can provide a microscopic illustration of the highly dynamic thrombus contracting process, a phenomenon that cannot be fully captured with current experimental techniques.

1. Nicholas Filla, Jixin Hou, He Li, Xianqiao Wang, “A Multiscale Framework for Modeling Fibrin Fiber Networks: Theory Development and Validation”, Journal of the Mechanics and Physics of Solids, 105392, 2023.
2. Nicholas Filla, Yiping Zhao, Xianqiao Wang, “Fibrin Fiber Deformation Mechanisms: Insights from Phenomenological Modeling to Molecular Details”, Biomechanics and Modeling in Mechanobiology, 1-19, 2023.
3. Nicholas Filla, Yiping Zhao, Xianqiao Wang, “A Tractable, Transferable, and Empirically Consistent Fibrous Material Model”, Polymers, 14(2): 4437, 2022.
4. Ning Liu, Liuyang Zhang, Matthew Becton and Xianqiao Wang. “Mechanism of Coupling Nanoparticle Stiffness with Shape for Endocytosis: From Rodlike Penetration to Wormlike Wriggling”, Journal of Physical Chemistry B, 124(49): 11145-11156, 2020.

Collaborators: Dr. Yiping Zhao (UGA); He Li (UGA); George Karniadakis (Brown U); Dr. Ning Liu (Tongji U)

Research Thrust III: Materials and Structures Design by AI

Kirigami-inspired designs hold great potential for the development of functional materials and devices, but predicting the morphological configuration of these structures under various loading conditions remains a challenge for traditional experimental and numerical methods. Here, we present a novel approach that utilizes machine learning algorithms to accurately predict the deformation and stress field of kirigami-inspired programmable active composites. To train our model, first we used a chemical corrosion algorithm to generate a dataset of kirigami-inspired imaging model accompanied by utilizing finite element simulations to obtain their deformation and stress fields as the ground truth, and then trained the machine learning model to offer robust predictions of the displacement and stress fields of the designated structures. Our results demonstrate the effectiveness of this approach in predicting the mechanical behavior of kirigami-inspired active structures, paving the way for the development of advanced and functional composite designs that are programmable and active.

Most existing ML approaches applied to the design of metamaterials are primarily based on a single property value with the assumption that the Poisson’s ratio of a material is stationary, neglecting the dynamic variability of Poisson’s ratio, termed deformation-dependent Poisson’s ratio, during the loading process a metamaterial other than conventional materials may experience. This project proposes a crystallographic symmetry-based methodology to build 2D metamaterials with complex but patterned topological structures, and then convert them into computational models suitable for molecular dynamics simulations. Then we employ an integrated approach, consisting of molecular dynamics simulations competent to generate and collect a large dataset for training/validation, and ML algorithms (CNN and cycle-GAN) able to predict the dynamic characteristics of Poisson’s ratio, as well as to offer the inverse design of a metamaterial structure based on a target quasi-continuous Poisson’s ratio-strain curve, to eventually unravel the underlying mechanism and design principles of 2D metamaterial structures with tunable Poisson’s ratio. The close match between the predefined Poisson’s ratio response and that from the generated structure validates the feasibility of the proposed ML model. Thanks to high efficiency and complete independence from prior knowledge, our proposed approach offers a novel and robust technique for the prediction and inverse design of metamaterial structures with tailored deformation-dependent Poisson’s ratio, an unprecedented property attractive in flexible electronics, soft robotics, nanodevices, etc.

1. Keke Tang, Yujie Xiang, Jie Tian, Jixin Hou, Xianyan Chen, Xianqiao Wang, and Zheng Zhong, “Machine learning-based morphological and mechanical prediction of kirigami-inspired active composites”, International Journal of Mechanical Sciences, 266, 108956, 2024.
2. Yuqing Cui, Yafei Xu, Donghai Han, Xingyu Wang, Xianqiao Wang, David S. Citrin, Liuyang Zhang, Ruqiang Yan, Xuefeng Chen, Asoke K. Nandi, “Hidden information extraction from ultra-layered structures through terahertz imaging down to ultra-low SNR”, Science Advances, 9(40), eadg8435, 2023.
3. Jie Tian, Keke Tang, Xianyan Chen, and Xianqiao Wang, “Machine Learning-based Prediction and Inverse Design of 2D Metamaterial Structures with Tunable Deformation-Dependent Poisson’s Ratio”, Nanoscale, 14(35): 12677-12691, 2022.
4. Zhen Zhang, Dai Han, Liuyang Zhang, Xianqiao Wang, Xuefeng Chen. “Adaptively reverse design of terahertz metamaterial for electromagnetically induced transparency with generative adversarial network”, Journal of Applied Physics, 130(3): 033101, 2021.

Collaborators: Dr. Kenan Song (UGA); Hongyue Sun (UGA); Keke Tang (Tongji U); Dr. Liuyang Zhang (Xi’an Jiaotong U)

Research Thrust VI: AI-empowered Engineering Training Program for Future Engineers

The demand for skilled engineers with expertise in artificial intelligence (AI) and machine learning (ML) has surged. To address this need, I plan to establish a state-of-the-art training program for students in the interdisciplinary engineering program. It aims to provide engineers with a distinctive focus on traditional mechanics, bioengineering, AI and ML applications, and collaboration skills. As the Principal Investigator (PI), I bring over years of hands-on experience of mechanical engineering and bioengineering research and student mentoring, positioning me uniquely to bridge the gap between mechanics, AI, ML, and interdisciplinary research and contribute significantly to the success of next generation of engineers.

1. Jie Tian, Jixin Hou, Zihao Wu, Peng Shu, Zhengliang Liu, Yujie Xiang, Beikang Gu, Nicholas Filla, Yiwei Li, Ning Liu, Xianyan Chen, Keke Tang, Tianming Liu, Xianqiao Wang, “Assessing Large Language Models in Mechanical Engineering Education: A Study on Mechanics-Focused Conceptual Understanding”, arXiv preprint arXiv:2401.1298, 2024.
2. Guoyu Lu, Sheng Li, Gengchen Mai, Jin Sun, Dajiang Zhu, Lilong Chai, Haijian Sun, Xianqiao Wang, Haixing Dai, Ninghao Liu, Rui Xu, Daniel Petti, Changying Li, Tianming Liu, “AGI for Agriculture”, arXiv preprint arXiv:2304.06136, 2023.

Collaborators: Dr. Tianming Liu (UGA); Xiaoming Zhai (UGA)