Automated Identification and Localization of Hematopoietic Stem Cells in 3D Intravital Microscopy Data

•A new tool allows automated 3D image analysis of HSCs and their niche•It performs automated segmentation of heterogeneous HSCs and bone marrow components•This tool identifies real HSCs and eliminates false-positive signals•3D distance measurements of HSC to the nearest osteoblast/bone are demonstrated

SummaryMeasuring three-dimensional (3D) localization of hematopoietic stem cells (HSCs) within the bone marrow microenvironment using intravital microscopy is a rapidly expanding research theme. This approach holds the key to understanding the detail of HSC-niche interactions, which are critical for appropriate stem cell function. Due to the complex tissue architecture of the bone marrow and to the progressive introduction of scattering and signal loss at increasing imaging depths, there is no ready-made software to handle efficient segmentation and unbiased analysis of the data. To address this, we developed an automated image analysis tool that simplifies and standardizes the biological interpretation of 3D HSC microenvironment images. The algorithm identifies HSCs and measures their localization relative to surrounding osteoblast cells and bone collagen. We demonstrate here the effectiveness, consistency, and accuracy of the proposed approach compared to current manual analysis and its wider applicability to analyze other 3D bone marrow components.