Build models to compare images of wildtype vs. disease cells, or to compare images of cell lines from different genotypes. Answer the question, do my cell models show a structural phenotype?
Build and deploy models of cellular disease to screen for drugs that can rescue a phenotype, or build models to look for structural changes due to drug toxicity.
Automatically segment regions of interest in your images. Then process the segmented images to create measurements of cell count, tissue prevalence, and other metrics.
Perform automated quality control. Train models that detect unwanted differentiated cells, out-of-focus images, or other types of artifacts.
Works with 2D images captured on any imaging system: brightfield, fluorescence, high-throughput imager, and more.
Train new models with minimal configuration, with no need to define image features a priori.
Provides objective, consistent, and unbiased results, with fast and scalable deployment.
Using PhenoLearn's automated deep learning image analysis platform, our team has been able to reproducibly identify subtle phenotypic differences in healthy and diseased cellular models. PhenoLearn’s artificial intelligence platform takes an unbiased approach to identify subtle subcellular and structural differences that are cumbersome to identify using traditional image analysis methods. Using PhenoLearn, our team has been able to identify a handful of protective druggable targets in cellular models and have been able translate our cell-based findings to preclinical models. Our expectation is that cell-based phenotypic screening will accelerate the rate of early stage drug and target discovery, leading to safer and more efficacious drugs.
PhenoLearn is an easy-to-use and intuitive software that has allowed me to segment and classify images using deep learning. Due to PhenoLearn’s plug-and-play structure, I was able to easily test a wide range of parameters to create meaningful, robust, and reliable models to quickly accomplish our company’s goals. As a visual learner, I also appreciated PhenoLearn’s visual representations of important metrics like accuracy and loss. This is indeed an excellent tool for biologists starting simple or complex deep learning projects.