Toward Robust Fully 3D Filopodium Segmentation and Tracking in Time-Lapse Fluorescence Microscopy

Investor logo
Investor logo
Investor logo

Warning

This publication doesn't include Institute of Computer Science. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

MAŠKA Martin NEČASOVÁ Tereza WIESNER David SOROKIN Dmitry PETERLÍK Igor ULMAN Vladimír SVOBODA David

Year of publication 2019
Type Article in Proceedings
Conference 26th IEEE International Conference on Image Processing
MU Faculty or unit

Faculty of Informatics

Citation
Web https://doi.org/10.1109/ICIP.2019.8803721
Doi http://dx.doi.org/10.1109/ICIP.2019.8803721
Keywords Benchmark dataset; synthetic image data; filopodium segmentation; filopodium tracking
Description Development, parameter tuning, and objective benchmarking of bioimage analysis workflows heavily rely on the availability of diverse bioimage datasets accompanied by reference annotations. In this paper, we present a new benchmark dataset, FiloData3D, designed for in-depth performance assessments of fully 3D filopodium segmentation and tracking algorithms that emerged recently in the field. It consists of 180 synthetic, fully annotated, 3D time-lapse sequences of single lung cancer cells, combining different cell shapes, signal-to-noise ratios, and anisotropy ratios, which are the well-known factors that influence the quality of segmentation and tracking results. Using FiloData3D, we show that the number of filopodia and their lengths extracted are significantly underestimated in the case of traditional 2D protocols that prevail in daily practice compared to fully 3D measurements, calling for a procedural change in filopodial analyses of 3D+t bioimage data.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.

More info