Deep-Class-CTCs - Deep-learning classification of dynamically flowing circulating tumors cells imaged by quantitative phase microscopy (DFM.AD001.403)
Area tematica
Scienze fisiche e tecnologie della materia
Area progettuale
Sensori multifunzionali e dispositivi elettronici (DFM.AD001)Struttura responsabile del progetto di ricerca
Istituto di Scienze Applicate e Sistemi Intelligenti "Eduardo Caianiello" (ISASI)
Responsabile di progetto
PIETRO FERRARO
Telefono: 0818675041
E-mail: pietro.ferraro@cnr.it
Abstract
The goal of this project is to develop dynamic deep-learning classifiers that can detect, analyze and monitor cancer cells in liquid blood samples. The blood samples will flow in a specialized micro-fluidic device that can rotate the cells during flow and provide an access to their perspective projections. The cells will be acquired by clinically-enabled interferometric imaging modules, developed together by both groups, providing the possibility to record the quantitative phase map projections of the blood and cancer cells in the full framerate of the camera used. Each of these topographic maps provides a great imaging contrast without cell staining, and also unique parameters that have not been attainable in previous attempts of imaging circulating tumor cells during flow, such as the cell volume, dry mass, and three-dimensional texture. For building the unified dataset of cancer cell maps, the optics-microfluidics system implementation and cell acquisition will be carried out in parallel in Italy and in Israel, where both groups have access to different types of cancer cells.
Data inizio attività
02/12/2021
Parole chiave
Digital holographic microscopy, Quantitative phase imaging, Biological cell imaging
Ultimo aggiornamento: 23/03/2025