Deep Label-Free Cell Imaging of Liquid Biopsies for Cancer Monitoring

Introduction

We will develop and commercialize an innovative device for diagnosis and monitoring of cancer in liquid biopsies based on a label-free interferometric phase microscopy (IPM) unit, coupled with dedicated real-time artificial intelligence (AI) for cell classification.

Innovative Approach

This device will materialize an innovative approach for the much-anticipated imaging flow cytometry, dramatically decreasing its costs and improving patient care by accurate monitoring of cancer in the clinical lab from a simple lab test (liquid biopsy).

Project Success Factors

The success of the project is dependent on four high-risk/high-gain aspects:

1. Building the first clinical IPM device.
2. Designing and manufacturing a disposable microfluidic device for imaging flow cytometry.
3. Obtaining high-enough acquisition and processing throughput in imaging flow cytometry of urine samples.
4. Training a deep neural network to detect cancer cells based on the information-deep label-free IPM images of cancer cells during flow.

Background

The proposed PoC project stems from my ongoing ERC StG project that focuses on the application of IPM for grading the metastatic potential of cancer cells, as a basic-science research tool.