Machine learning applications in cell image analysis

Abstract

Machine learning (ML) refers to a set of automatic pattern recognition methods that have been successfully applied across various problem domains, including biomedical image analysis. This review focuses on ML applications for image analysis in light microscopy experiments with typical tasks of segmenting and tracking individual cells, and modelling of reconstructed lineage trees. After describing a typical image analysis pipeline and highlighting challenges of automatic analysis (for example, variability in cell morphology, tracking in presence of clutters) this review gives a brief historical outlook of ML, followed by basic concepts and definitions required for understanding examples. This article then presents several example applications at various image processing stages, including the use of supervised learning methods for improving cell segmentation, and the application of active learning for tracking. The review concludes with remarks on parameter setting and future directions.

Advanced microscopy and imaging techniques in immunology and cell biology

“…These systems are providing biologists with an unprecedented level of resolution, but this does come with drawbacks. The most pressing issue is how we analyse and manage these huge data sets 4 , 5 . Thus, the field is currently at a crossroad where we desperately need to recruit computer scientists to specifically focus on developing novel, standardised methods for data analysis 5 .…”

“…Techniques such as in vitro time‐lapse tracking, 1 IVM, 2 , 3 3D imaging 4 and super‐resolution microscopy techniques 1 (all discussed in this Special Feature) are leading to an explosion of data similar to genomic data over the past 2 decades. This microscopy data requires bespoke, rigorously designed methods to extract quantitative information for in depth statistical interrogation 5 . Therefore, the time for investment in these careers and integration into research teams at the faculty level is now.…”

“…Approaches currently in widespread clinical use include Positron Emission Tomography and Magnetic Resonance Imaging. However, each day new uses for the technology discussed in this Special Feature 1 , 2 , 3 , 4 , 5 , 6 are being modified for exciting future clinical application. For example, integration of IVM style technology in traditional endoscopes that is advancing the analysis and diagnosis of cancer.…”

“…4,5 Thus, the field is currently at a crossroad where we desperately need to recruit computer scientists to specifically focus on developing novel, standardised methods for data analysis. 5 To address this, the onus is on biologists and institutions that rely on advanced microscopy to provide clear career development pathways so that we can recruit and retain the very best computer scientists. My personal belief is that the field of image analysis is in a stage of metamorphosis similar to bioinformatics in the early 2000s.…”

“…This microscopy data requires bespoke, rigorously designed methods to extract quantitative information for in depth statistical interrogation. 5 Therefore, the time for investment in these careers and integration into research teams at the faculty level is now.…”