The article introduces AxonTracker-3D, an interactive 3D axon tracking and labeling tool designed to reconstruct the neuromuscular junction connectome. The connectome, which describes the full connectivity of nervous system neuronal components, is crucial for understanding the organizational and developmental properties of the mammalian central nervous system (CNS). The tool addresses the challenge of automatically tracking and reconstructing axons from microscopy images in 3D, which is essential for studying synapse elimination during development and other phenomena in the CNS. The authors highlight the limitations of existing methods, such as the need for manual intervention, low contrast, and orientation changes, which make it difficult to track complex axon structures. AxonTracker-3D overcomes these challenges by incorporating a real-time tracking approach based on diffusion, reducing the time required for processing large volumes of axon images from months to minutes on personal computers. The method involves preprocessing the raw image datasets to reduce noise and extract centerlines of axons using a combination of gradient vector flow (GVF) and object orientation vectors. The tracking process starts with manually chosen initial points and continues with real-time updates and user intervention for correction. The segmented axons are then stitched together to form a complete connectome. The article also discusses the future improvements planned for the software, including the detection of branching points and the implementation of multi-threading for faster execution. The authors emphasize the potential of the quantitative data generated from the connectome studies to address biological questions about axon branching, spatial distribution, and connectivity.The article introduces AxonTracker-3D, an interactive 3D axon tracking and labeling tool designed to reconstruct the neuromuscular junction connectome. The connectome, which describes the full connectivity of nervous system neuronal components, is crucial for understanding the organizational and developmental properties of the mammalian central nervous system (CNS). The tool addresses the challenge of automatically tracking and reconstructing axons from microscopy images in 3D, which is essential for studying synapse elimination during development and other phenomena in the CNS. The authors highlight the limitations of existing methods, such as the need for manual intervention, low contrast, and orientation changes, which make it difficult to track complex axon structures. AxonTracker-3D overcomes these challenges by incorporating a real-time tracking approach based on diffusion, reducing the time required for processing large volumes of axon images from months to minutes on personal computers. The method involves preprocessing the raw image datasets to reduce noise and extract centerlines of axons using a combination of gradient vector flow (GVF) and object orientation vectors. The tracking process starts with manually chosen initial points and continues with real-time updates and user intervention for correction. The segmented axons are then stitched together to form a complete connectome. The article also discusses the future improvements planned for the software, including the detection of branching points and the implementation of multi-threading for faster execution. The authors emphasize the potential of the quantitative data generated from the connectome studies to address biological questions about axon branching, spatial distribution, and connectivity.