Deep learning classification for macrophage subtypes through cell migratory pattern analysis

• DOI: 10.3389/fcell.2024.1259037
• Publication Date: 2024-02-07T04:55:58Z
• Authors: Manasa Kesapragada; Yao-Hui Sun; Ksenia Zlobina; Cynthia Recendez; Daniel Fregoso; Hsin-Ya Yang; Elham Aslankoohi; Rivkah Isseroff; Marco Rolandi; Min Zhao; Marcella Gomez
• Keywords: Biomedical and clinical sciences; QH301-705.5; macrophage polarization; deep learning for classification of macrophage subtypes; Biological Sciences; correlation between cell shape and trajectories; Biological sciences; Cell and Developmental Biology; 03 medical and health sciences; 0302 clinical medicine; classification of macrophage subtypes using migration patterns; analysis of macrophage trajectory patterns; Biochemistry and Cell Biology; Biology (General)

Macrophages can exhibit pro-inflammatory or pro-reparatory functions, contingent upon their specific activation state. This dynamic behavior empowers macrophages to engage in immune reactions and contribute tissue homeostasis. Understanding the intricate interplay between macrophage motility status provides valuable insights into complex mechanisms that govern diverse functions. In a recent study, we developed classification method based on morphology, which demonstrated movement characteristics, including speed displacement, serve as distinguishing factors for subtypes. this develop deep learning model explore potential of classifying subtypes solely raw trajectory patterns. The relies time series x-y coordinates, well distance traveled net displacement. We begin by investigating migratory patterns gain deeper understanding behavior. Although analysis does not directly inform model, it serves highlight distinct dynamics exhibited different subtypes, cannot be easily captured finite set metrics. Our study uses cell trajectories classify three subtypes: M0, M1, M2. advancement holds promising implications future, suggests possibility identifying without relying shape analysis. Consequently, could potentially eliminate necessity high-quality imaging techniques provide more robust methods analyzing inherently blurry images.