A novel universal algorithm for filament network tracing and cytoskeleton analysis

The rapid development of advanced microscopy techniques over recent decades has significantly increased the quality of imaging and our understanding of subcellular structures, such as the organization of the filaments of the cytoskeleton using fluorescence and electron microscopy. However, these recent improvements in imaging techniques have not been matched by similar development of techniques for computational analysis of the images of filament networks that can now be obtained. Hence, for a wide range of applications, reliable computational analysis of such two-dimensional methods remains challenging. Here, we present a new algorithm for tracing of filament networks. This software can extract many important parameters from grayscale images of filament networks, including the mesh hole size, and filament length and connectivity (also known as Coordination Number). In addition, the method allows sub-networks to be distinguished in two-dimensional images using intensity thresholding. We show that the algorithm can be used to analyze images of cytoskeleton networks obtained using different advanced microscopy methods. We have thus developed a new improved method for computational analysis of two-dimensional images of filamentous networks that has wide applications for existing imaging techniques. The algorithm is available as open-source software.

Automated summary

Advancements in advanced microscopy techniques have enhanced imaging quality and understanding of subcellular structures, but computational analysis techniques have not progressed at the same pace. A new algorithm for tracing filament networks has been developed, capable of extracting important parameters and distinguishing sub-networks in two-dimensional images. This algorithm can be widely applied in analyzing images obtained from different advanced microscopy methods.