Nonlinear Shape Regression For Filtering Segmentation Results From Calcium Imaging

A shape filter is presented to repair segmentation results obtained in calcium imaging of neurons in vivo. This post-segmentation algorithm can automatically smooth the shapes obtained from a preliminary segmentation, while precluding the cases where two neurons are counted as one combined component. The shape filter is realized using a square-root velocity to project the shapes on a shape manifold in which distances between shapes are based on elastic changes. Two data-driven weighting methods are proposed to achieve a trade-off between shape smoothness and consistency with the data. Intuitive comparisons of proposed methods via projection onto Cartesian maps demonstrate the smoothing ability of the shape filter. Quantitative measures also prove the superiority of our methods over models that do not employ any weighting criterion.
Comment: v1: Submitted as a conference contribution to ICIP 2018 (pre-print version Feb. 14, 2018); v2: Accepted to be published in 2018 IEEE International Conference on Image Processing, Oct 7-10, 2018, Athens, Greece. IEEE Copyright notice added. Minor changes for camera-ready version (updated Jun. 1, 2018)