A method is provided for non-invasively predicting characteristics of one or more cells and cell derivatives. The method includes training a machine learning model using at least one of a plurality of training cell images representing a plurality of cells and data identifying characteristics for the plurality of cells. The method further includes receiving at least one test cell image representing at least one test cell being evaluated, the at least one test cell image being acquired non-invasively and based on absorbance as an absolute measure of light, and providing the at least one test cell image to the trained machine learning model. Using machine learning based on the trained machine learning model, characteristics of the at least one test cell are predicted. The method further includes generating, by the trained machine learning model, release criteria for clinical preparations of cells based on the predicted characteristics of the at least one test cell.
We have invented a non-invasive process for assessing the quality of manufactured batches of a tissue engineered retinal pigment epithelium (RPE) that are intended for human use as a therapeutic device for treating eye disorders. Manufactured batches are imaged in a brightfield microscope using a quantitative absorbance imaging routine. The samples are imaged in their entirety using an automated microsope resulting in a "big data" set (millions of images, terabytes of data). The images are analyzed by a machine learning algorithm to assess RPE quality based on the patterns of absorbance, which are related to melanin expression by the RPE. Native RPE express melanin in a particular pattern which is challenging to describe with a mathematical model but which can be effectively quantified by artificial intelligence using millions of parameters.