Machine Learning-Based Quantification of Motility Parameters in Individual Leptospira Cells.

Leptospira spp. are characterized by highly active motility, exhibiting swimming in liquid environments and crawling on solid surfaces, including host cells. Because motility of pathogenic Leptospira spp. is closely associated with successful infection, a deeper understanding of the underlying mechanism of motility and cellular behavior in response to environmental changes is essential for advancing countermeasures against leptospirosis. However, the precise role of motility as a virulence factor remains incompletely understood. One major obstacle has been the technical difficulty of quantitative analyzing complex motility patterns in a high-throughput manner. This chapter describes methodologies for automatic measurement of motility parameters, such as swimming and crawling speeds, using machine learning-based recognition and tracking of individual leptospiral cells.