Prediction of Fluorescent Ki67 Staining in 3D Tumor Spheroids

Abstract 3D cell culture models are important tools for the development and testing of new therapeutics. In combination with immunoassays confocal microscopy, crucial information like morphological or metabolic changes can be examined during drug testing. However, a common limitation immunostainings is number dyes that imaged simultaneously, as overlaps in spectral profiles different may result cross talk. We therefore present deep learning method, able to predict fluorescent stainings specific antigens on basis nuclei staining. Using proliferation marker Ki67, we showed presented model was Ki67 staining strong correlation real signal. Additional analysis showed, not relying signal This approach, based subsequent prediction target antigen, could reduce parallel antigen stains minimum incompatible panels circumvented future.