Deepcell Uses AI to Phenotype and Enrich Live Single Cancer Cells, Scheduled to Present Poster at AGBT 2021
Deepcell presents AI-powered approach for morphology-based cell characterization, identification and sorting of unlabeled cells with unprecedented resolution and accuracy
MOUNTAIN VIEW, Calif. – February 23, 2021 − Deepcell, a life science company pioneering AI-powered cell classification and isolation for cell biology and translational research, will present a poster at the AGBT 2021 virtual conference. The poster features new data generated on the Deepcell platform, showing that the technology can quantitatively phenotype and enrich live single cancer cells. Deepcell’s unique AI-powered technology transforms cell morphology into a precise, reproducible and unbiased analyte that enables high accuracy cell classification while maintaining cell viability. Applying deep learning capabilities to cell biology, the Deepcell platform combines high-resolution imaging of cells in flow with real-time cell classification and sorting using cell morphology as the single analyte.
To date, Deepcell’s AI-powered classifier has been trained on tens of millions of images of multiple types of cells, as part of the company’s fast-growing Cell Morphology Atlas. As a result, it has accurately discriminated among immune cell subtypes, cells from various cancers and stromal cells against a background of blood cells.
This label-free, target-agnostic approach overcomes some of the limitations of cell surface marker-based classification and enrichment, including the limited number of available markers and channels for detection, prior knowledge or guesswork required to select surface proteins, and availability of protein-specific antibodies.
To go a step further, the company has demonstrated that unlabeled sorted cells from the Deepcell platform retain viability and are suitable for single-cell multi-omic analysis, thereby enabling integration of quantitative cell morphology with multi-omic approaches for an unprecedented understanding of the cell.
“Deepcell’s data illustrates that deep learning can achieve high classification accuracies, reveal new ways to precisely characterize and phenotype cells, and enable the label-free purification of cells of interest for further molecular analysis,” said Mahyar Salek, CTO of Deepcell. “This technology offers a new tool to biologists and research scientists in academia, translational research institutes and the pharma/biotech industry to contextualize and derive insights from cell morphology data.”
For more information about Deepcell, please go to www.deepcell.com.
Deepcell is helping to advance precision medicine by combining advances in AI, cell classification and capture, and single-cell analysis to deliver novel insights through an unprecedented view of cell biology. Spun out of Stanford University in 2017, the company has created unique, microfluidics-based technology that uses continuously learning AI to classify cells based on detailed visual features and sort them without inherent bias. The Deepcell platform maintains cell viability for downstream single-cell analysis and can be used to isolate virtually any type of cell, even those occurring at frequencies as low as one in a billion. The technology will initially be available as a service for use in translational research as well as diagnostics and therapeutic development. Deepcell is privately held and based in Mountain View, CA. For more information, please visit deepcellbio.com.
# # #