Search

Segmentation is required to quantify cellular structures in microscopic images. This typically requires their fluorescent labeling. Convolutional neural networks (CNNs) can detect these structures also in only transmitted light images. This eliminates the need for transgenic or dye fluorescent labeling, frees up imaging channels, reduces phototoxicity and speeds up imaging. However, this approach currently requires optimized experimental conditions and computational specialists. Here, we introduce "aiSEGcell" a user-friendly CNN-based software to segment nuclei and cells in bright field images. We extensively evaluated it for nucleus segmentation in different primary cell types in 2D cultures from different imaging modalities in hand-curated published and novel imaging data sets. We provide this curated ground-truth data with 1.1 million nuclei in 20,000 images. aiSEGcell accurately segments nuclei from even challenging bright field images, very similar to manual segmentation. It retains biologically relevant information, e.g. for demanding quantification of noisy biosensors reporting signaling pathway activity dynamics. aiSEGcell is readily adaptable to new use cases with only 32 images required for retraining. aiSEGcell is accessible through both a command line, and a napari graphical user interface. It is agnostic to computational environments and does not require user expert coding experience. Author summary: Fluorescence microscopy is the most widely used method to monitor cellular structures in space and time. Fluorescently labeling cellular structures is typically required to localize ("segment") them in electronic images for subsequent quantification. Deep learning approaches can detect these structures also in only bright field images. This eliminates the need for a fluorescent label, frees up imaging channels, speeds up imaging, and reduces the harmful effects of exposing cells to high intensity light. However, label free segmentation currently requires optimized experimental conditions and computational specialists. Therefore, we developed "aiSEGcell" a user-friendly deep learning-based software to segment nuclei and cells in only bright field images. We extensively evaluated aiSEGcell on different common experimental conditions and showed that biologically even sensitive relevant information is retained. Furthermore, we demonstrated that aiSEGcell is adaptable by retraining to new applications with very little required data. We make it accessible for users with no required expert coding experience in a wide range of computational environments. Finally, we openly share our very large imaging data sets to further the development of other segmentation approaches. [ABSTRACT FROM AUTHOR]