Title: Automated detection and enumeration of bead encapsulation in microfluidic droplets based on deep learning

Authors: Hoang Anh Phan; Loc Quang Do; Thanh Tung Bui; Thang Nguyen Van; Hoang Hai Nguyen; Trinh Chu Duc

Addresses: Faculty of Electronics and Telecommunication, MEMs and Microsystem Department, University of Engineering and Technology, Vietnam National University, Hanoi 144, Xuan Thuy, Cau Giay, Hanoi, Vietnam ' Faculty of Physics, Radiophysics Department, Hanoi University of Science, Vietnam National University, Hanoi, 334, Nguyen Trai, Hanoi, Vietnam ' Faculty of Electronics and Telecommunication, MEMs and Microsystem Department, University of Engineering and Technology, Vietnam National University, Hanoi, 144, Xuan Thuy, Cau Giay, Hanoi, Vietnam ' Faculty of Electronics and Telecommunication, Department of Electronics and Computer Engineering, University of Engineering and Technology, Vietnam National University, Hanoi, 144, Xuan Thuy, Cau Giay, Hanoi, Vietnam ' Vietnam National University, Hanoi, 144, Xuan Thuy, Cau Giay, Hanoi, Vietnam ' University of Engineering and Technology, Vietnam National University, Hanoi 144, Xuan Thuy, Cau Giay, Hanoi, Vietnam

Abstract: The encapsulation of cells within droplets is a crucial aspect of various cell analysis applications. Current research has primarily focused on detecting and identifying cell types or cell counts within droplets using object detection models in bright-field images. However, there is a lack of research into statistical approaches that can improve the reliability of droplet classification systems by accurately quantifying the number of cells present in each droplet. In this study, a microfluidic droplet generation system was developed and implemented, coupled with optical devices to capture images of encapsulated beads within the droplet. To achieve high accuracy and reliability of the system, a statistical method was proposed to detect and enumerate beads within a droplet. An experimental investigation was conducted on the droplet formation channel to assess its capability to beads encapsulation. The preprocessing procedures for the images obtained from the optical system's video recording include identifying and cropping droplets to create a dataset. This dataset is then used to train and validate the deep learning model to detect droplets and beads in microfluidic devices. Precision and recall were used as metrics in both the validation and test sets, contributing to a notably high mean average precision (mAP) score, signifying the accuracy of the droplet detector. The experimental results show the potential significance of the proposed method in enhancing the reliability of single-cell sorting systems. This capability paves the way for isolating rare cell populations, facilitating their subsequent in-depth analysis.

Keywords: single-cell sorting; deep learning; microfluidic; droplet; bead encapsulation.

DOI: 10.1504/IJNT.2024.146467

International Journal of Nanotechnology, 2024 Vol.21 No.7/8/9/10/11/12, pp.609 - 621

Published online: 30 May 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article