Adaptive-Resolution Multi-Orientation Analysis of Complex Filamentous Network Images

Microscopy images of cytoskeletal, nucleoskeletal, and other filamentous structures contain complex junctions where multiple filaments with distinct orientations overlap, yet state-of-the-art software generally uses single orientation analysis to segment these structures. We describe an image analysis approach to simultaneously segment both filamentous structures and their intersections in microscopy images, based on analytically resolving coincident multiple orientations upon image filtering in a manner that balances orientation resolution and spatial localization.