|
Abstract:
|
Several recent cell imaging technologies such as Cyclic Immunoflourescence are able to capture hundreds of cell image features, often by applying multiple stains to each cell, allowing researchers to explore cellular phenotypic responses in unprecedented detail. However, high throughput cell imaging also presents several challenges, including a highly heterogeneous set of image features, spatial biases, segmentation errors, variation in stain intensity, and degradation due to multiple rounds of staining, among others. We discuss the design and analysis of cell imaging studies, highlighting in particular the importance of multiple forms of replication to help address these challenges. We also discuss the subsequent analysis of cell image data, discussing methods to integrate information from multiple image features, as well as again highlighting the critical role of multiple forms of replication. We then present a complete analysis pipeline, and apply it to microenvironment microarray and cyclic immunoflourescence studies.
|
