3D shape analysis using overcomplete spherical wavelets: Application to BLEB detection in cell biology

Abstract

Amoeboid cell motility is characterised by the emission of protrusions at the cellular surface (also known as “blebs”). Detection and counting of these protrusions is a crucial step towards the understanding of the deformation and motility machinery. We propose an automated technique to detect protrusions at the surface of cells observed in 3D fluorescence microscopy using over-complete spherical wavelets. The framework permits intuitive manipulation of wavelets on the sphere, thanks to a straightforward analogy with traditional wavelets on the plane. We illustrate detection results on a real data set of protruding cells, indicating the reliability of the method. Moreover, the flexibility of the approach makes it easily amenable to other shape analysis problems.