cnn-transfer-poster.pdf (976.27 kB)
Transfer learning with deep convolutional neural networks for classifying cellular morphological changes
poster
posted on 2019-08-29, 13:10 authored by Alexander Kensert, Philip HarrisonPhilip Harrison, Ola SpjuthOla SpjuthIn this study we applied pre-trained CNNs to predict cell mechanisms of action (MoAs) in response to chemical perturbations for two cell profiling datasets from the Broad Bioimage Benchmark Collection (bbbc) and obtained higher predictive accuracy than previously reported, between 95 and 97%.