Periodic Reporting for period 1 - HoloStain (Imaging flow cytometry using holographic virtual staining)
Reporting period: 2019-07-01 to 2021-06-30
Imaging flow cytometry (IFC) has a great potential for becoming a diagnosis tool, allowing imaging, analyzing and processing a large number of cells from body fluids, which is highly relevant for blood analysis and cancer detection. However, at the moment, IFC measures cell structures by staining different cell parts on different locations of the camera sensor, dictating a very complicated optical setup as well as expensive cameras and processing units. Current IFC remains a highly inaccessible technology, due to its very high cost, lack of accuracy, required operator expertise, and lack of objectiveness. The global flow cytometry market size is estimated as USD 3.29 Billion in 2017, and projected to reach USD 4.79 Billion by 2022.
In this project, we developed a new type of IFC based on holographic imaging, which can record live cells during flow without staining in a quantitative manner, and a deep-learning framework that can covert the quantitative imaging data and make it look as if the cells have been stained but without using chemical staining. These developments significantly decreases IFC machine sizes, making them more accessible for direct clinical use.
Research and development activities in the project included: building a robust deep neural network that provided correct virtually stained images of cells, designing and building a portable module suitable for IFC, speeding up digital processing implementations, preliminary experiments with IFC of white blood cells and cell lines. A research paper was published in PNAS, one of the best scientific journals. Pre-commercialization activities in the project included: completing an in-depth market analysis as well as analyzing the strengths, weaknesses, threats and opportunities of the proposed technology, mapping the users and global manufacturers of IFC instruments and reagents, with an elaborated analysis on the future directions, preparing a marketing requirement document for the product realization pathways. This document expressed the customer's requirements and needs for the product, and preparing an investor presentation. The project is now under due-diligence of three entrepreneurs, with the goal of forming a start-up company and continuing the commercialization path. We also plan submitting to the EIC Transition Grant to move this project forward.
In this project, we developed a new type of IFC based on holographic imaging, which can record live cells during flow without staining in a quantitative manner, and a deep-learning framework that can covert the quantitative imaging data and make it look as if the cells have been stained but without using chemical staining. These developments significantly decreases IFC machine sizes, making them more accessible for direct clinical use.
Research and development activities in the project included: building a robust deep neural network that provided correct virtually stained images of cells, designing and building a portable module suitable for IFC, speeding up digital processing implementations, preliminary experiments with IFC of white blood cells and cell lines. A research paper was published in PNAS, one of the best scientific journals. Pre-commercialization activities in the project included: completing an in-depth market analysis as well as analyzing the strengths, weaknesses, threats and opportunities of the proposed technology, mapping the users and global manufacturers of IFC instruments and reagents, with an elaborated analysis on the future directions, preparing a marketing requirement document for the product realization pathways. This document expressed the customer's requirements and needs for the product, and preparing an investor presentation. The project is now under due-diligence of three entrepreneurs, with the goal of forming a start-up company and continuing the commercialization path. We also plan submitting to the EIC Transition Grant to move this project forward.