Information

Related Research Units

Vascular Biology Program Research

Research Overview

Recent advances in microscopy, live-cell imaging, spatial omics, and single-cell technologies are transforming cell biology into a data-rich and increasingly dynamic science. These technologies now reveal levels of cellular heterogeneity that were previously inaccessible, creating a major “big data” challenge for biomedical research. Artificial intelligence (AI) and machine learning provide powerful tools for extracting meaningful patterns from high-dimensional and heterogeneous biological datasets. However, many current AI applications in cell biology still focus primarily on static data, such as endpoint microscopy images or single-cell RNA-seq profiles, while the rich dynamic information contained in high-resolution live-cell movies remains underutilized. My research addresses this gap by developing AI platforms to discover, quantify, and interpret dynamic cellular phenotypes from live-cell imaging data. We focus on cellular morphodynamics, motility, cell-cell interactions, and temporal behavioral states that are difficult or impossible to detect by human observation alone. By combining deep learning, self-supervised representation learning, heterogeneity-preserving feature selection, and multimodal integration with single-cell and spatial transcriptomic data, our goal is to connect dynamic cell behavior with molecular cell states and disease mechanisms. 

Using these approaches, my laboratory studies phenotypic heterogeneity in cancer, airway epithelial differentiation, and other disease-relevant cellular systems. We aim to identify hidden dynamic cell states, understand how they contribute to disease progression and treatment response, and develop AI-guided strategies for precision diagnosis and therapeutics. Ultimately, our work seeks to establish live-cell morphodynamics as a new quantitative dimension of cell biology and precision medicine.

Research Background

Click here to be redirected to the Kwonmoo Lee Laboratory

Dr. Kwonmoo Lee received his graduate training in Marc Kirschner’s laboratory at Harvard Medical School while completing his Ph.D. in Physics at MIT. He then pursued postdoctoral training as an NIH postdoctoral fellow in Gaudenz Danuser’s laboratory at Harvard Medical School, where he developed expertise in quantitative cell biology, live-cell imaging, and computational analysis of cellular dynamics. He later served as an Assistant Professor in the Department of Biomedical Engineering at Worcester Polytechnic Institute before joining the Vascular Biology Program at Boston Children’s Hospital and the Department of Surgery at Harvard Medical School in 2020. 

Dr. Lee’s research focuses on developing artificial intelligence and machine learning methods to analyze cellular heterogeneity in cell biology, mechanobiology, and disease. In particular, his laboratory studies how cellular and subcellular morphodynamic heterogeneity, cell motility, and dynamic cell-state transitions regulate cellular behavior in healthy and pathological conditions. By integrating live-cell imaging, deep learning, and single-cell/spatial omics, his work aims to uncover hidden dynamic phenotypes that contribute to angiogenesis, cancer progression, treatment response, and other disease-relevant biological processes.

Publications

  1. Diffraction-informed deep learning for molecular-specific holograms of breast cancer cells. APL Bioeng. 2025 Sep; 9(3):036107. View Abstract
  2. Survivin modulates stiffness-induced vascular smooth muscle cell motility. APL Bioeng. 2025 Jun; 9(2):026120. View Abstract
  3. Heterogeneity-preserving discriminative feature selection for disease-specific subtype discovery. Nat Commun. 2025 Apr 16; 16(1):3593. View Abstract
  4. Fine-grained forecasting of COVID-19 trends at the county level in the United States. NPJ Digit Med. 2025 Apr 11; 8(1):204. View Abstract
  5. Breast Cancer-Derived Extracellular Vesicles Modulate the Cytoplasmic and Cytoskeletal Dynamics of Blood-Brain Barrier Endothelial Cells. J Extracell Vesicles. 2025 Jan; 14(1):e70038. View Abstract
  6. Survivin modulates stiffness-induced vascular smooth muscle cell motility. bioRxiv. 2024 Dec 12. View Abstract
  7. Interpretable Fine-Grained Phenotypes of Subcellular Dynamics via Unsupervised Deep Learning. Adv Sci (Weinh). 2024 11; 11(41):e2403547. View Abstract
  8. Deep learning-based automated analysis of NK cell cytotoxicity in single cancer cell arrays. BioChip Journal. 2024. View Abstract
  9. Survivin as a mediator of stiffness-induced cell cycle progression and proliferation of vascular smooth muscle cells. APL Bioeng. 2023 Dec; 7(4):046108. View Abstract
  10. Survivin regulates intracellular stiffness and extracellular matrix production in vascular smooth muscle cells. APL Bioeng. 2023 Dec; 7(4):046104. View Abstract
  11. Unsupervised Contour Tracking of Live Cells by Mechanical and Cycle Consistency Losses. Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2023 Jun; 2023:227-236. View Abstract
  12. Screening adequacy of unstained thyroid fine needle aspiration samples using a deep learning-based classifier. Sci Rep. 2023 08 19; 13(1):13525. View Abstract
  13. Deep cross-modal feature learning applied to predict acutely decompensated heart failure using in-home collected electrocardiography and transthoracic bioimpedance. Artif Intell Med. 2023 06; 140:102548. View Abstract
  14. Protocol for live cell image segmentation to profile cellular morphodynamics using MARS-Net. STAR Protoc. 2022 09 16; 3(3):101469. View Abstract
  15. A machine learning pipeline revealing heterogeneous responses to drug perturbations on vascular smooth muscle cell spheroid morphology and formation. Sci Rep. 2021 12 02; 11(1):23285. View Abstract
  16. Increased prevalence of indoor Aspergillus and Penicillium species is associated with indoor flooding and coastal proximity: a case study of 28 moldy buildings. Environ Sci Process Impacts. 2021 Nov 17; 23(11):1681-1687. View Abstract
  17. A deep learning-based segmentation pipeline for profiling cellular morphodynamics using multiple types of live cell microscopy. Cell Rep Methods. 2021 11 22; 1(7). View Abstract
  18. Emerging machine learning approaches to phenotyping cellular motility and morphodynamics. Phys Biol. 2021 06 17; 18(4). View Abstract
  19. Mechanosensitive expression of lamellipodin promotes intracellular stiffness, cyclin expression and cell proliferation. J Cell Sci. 2021 06 15; 134(12). View Abstract
  20. Mechanosensitive expression of lamellipodin promotes intracellular stiffness, cyclin expression, and cell proliferation. J Cell Sci. 2021 May 25. View Abstract
  21. Deep transfer learning-based hologram classification for molecular diagnostics. Sci Rep. 2018 11 19; 8(1):17003. View Abstract
  22. Deconvolution of subcellular protrusion heterogeneity and the underlying actin regulator dynamics from live cell imaging. Nat Commun. 2018 04 27; 9(1):1688. View Abstract
  23. Functional hierarchy of redundant actin assembly factors revealed by fine-grained registration of intrinsic image fluctuations. Cell Syst. 2015 Jul 29; 1(1):37-50. View Abstract
  24. Self-assembly of filopodia-like structures on supported lipid bilayers. Science. 2010 Sep 10; 329(5997):1341-5. View Abstract
  25. A Stochastic Model of Conductance Transitions in Voltage-Gated IonChannels. J Biol Phys. 2002 Jun; 28(2):279-88. View Abstract
  26. Effects of nonequilibrium fluctuations on ionic transport through biomembranes. Phys Rev E Stat Phys Plasmas Fluids Relat Interdiscip Topics. 1999 Oct; 60(4 Pt B):4681-6. View Abstract

Contact Kwonmoo Lee

Email:
Print Profile