An objective comparison of cell-tracking algorithms
Ulman, Vladimír and Maška, Martin and Magnusson, Klas E. G. and Ronneberger, Olaf and Haubold, Carsten and Harder, Nathalie and Matula, Pavel and Matula, Petr and Svoboda, David and Radojevic, Miroslav and Smal, Ihor and Rohr, Karl and Jaldén, Joakim and Blau, Helen M. and Dzyubachyk, Oleh and Lelieveldt, Boudewijn and Xiao, Pengdong and Li, Yuexiang and Cho, Siu-Yeung and Dufour, Alexandre C. and Jean-Christophe Olivo-Marin, Constantino C. Reyes-Aldasoro, Jose A. Solis-Lemus, Robert Bensch, and Brox, Thomas and Stegmaier, Johannes and Mikut, Ralf and Wolf, Steffen and Hamprecht, Fred A. and Esteves, Tiago and Quelhas, Pedro and Demirel, Ömer and Malmström, Lars and Jug, Florian and Tomancak, Pavel and Meijering, Erik and Muñoz-Barrutia, Arrate and Kozubek, Michal and Ortiz-de-Solorzano, CarlosAbstract
We present a combined report on the results of three editions of the Cell Tracking Challenge, an ongoing initiative aimed at promoting the development and objective evaluation of cell segmentation and tracking algorithms. With 21 participating algorithms and a data repository consisting of 13 data sets from various microscopy modalities, the challenge displays today’s state-of-the-art methodology in the field. We analyzed the challenge results using performance measures for segmentation and tracking that rank all participating methods. We also analyzed the performance of all of the algorithms in terms of biological measures and practical usability. Although some methods scored high in all technical aspects, none obtained fully correct solutions. We found that methods that either take prior information into account using learning strategies or analyze cells in a global spatiotemporal video context performed better than other methods under the segmentation and tracking scenarios included in the challenge.