DeepIFC: Virtual fluorescent labeling of blood cells in imaging flow cytometry data with deep learning

Abstract

Imaging flow cytometry (IFC) combines flow cytometry with microscopy, allowing rapid characterization of cellular and molecular properties via high‐throughput single‐cell fluorescent imaging. However, fluorescent labeling is costly and time‐consuming. We present a computational method called DeepIFC based on the Inception U‐Net neural network architecture, able to generate fluorescent marker images and learn morphological features from IFC brightfield and darkfield images. Furthermore, the DeepIFC workflow identifies cell types from the generated fluorescent images and visualizes the single‐cell features generated in a 2D space. We demonstrate that rarer cell types are predicted well when a balanced data set is used to train the model, and the model is able to recognize red blood cells not seen during model training as a distinct entity. In summary, DeepIFC allows accurate cell reconstruction, typing and recognition of unseen cell types from brightfield and darkfield images via virtual fluorescent labeling.

Morphological single-cell analysis of peripheral blood mononuclear cells from 390 healthy blood donors with Blood Cell Painting

“…We envision our BCP data to facilitate self-supervised training of very large encoder models (Bommasani et al 2021, Pfaendler et al 2023 for modeling morphological and functional characteristics of PBMCs, leading to the discovery of novel genotype-phenotype associations and blood cell trait determinants. Another exciting direction is training label-free models for inexpensive morphological profiling (Cross-Zamirski et al 2022, Timonen et al 2023 with large BCP datasets. Our results will pave the way for understanding the causes and consequences of hematological phenotypes in disease and health by contributing a large-scale morphological characterization of immune cells in healthy individuals.…”