Application of Hydrodynamic Focusing in Flow Cytometry and Solve Partial Differential Equations using Machine Learning

Aug 5, 2026

Upload by Quinn

Microfluidic flow cytometry offers portability and reduces sample and sheath-fluid consumption, yet device design has often depended on empirical trial-and-error rather than a precise, parameter-driven understanding of performance. This thesis uses computational fluid dynamics simulations to quantify how sheath-fluid velocity and related conditions influence the focused-stream width. It shows that increasing the velocity ratio can reduce stream width in the channel before laser interrogation past a threshold. It further frames machine-learning approaches to efficiently solve partial differential equations such as momentum and energy conservation equations.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages97

Reading Time4 h 4 min

AccessFree to download

Abstract

Declaration

Acknowledgements

Table of Contents

List of Figures

List of Tables

List of Acronyms

List of Glossary

1 Introduction

1.1 Fluid Dynamic Focusing Background

1.2 Machine Learning Background

1.3 Motivation for the present research work

2 Literature Review

2.1 Flow Cytometer Geometry Literature Review

2.2 Meshing Literature Reviews

2.3 Velocity Coupling Ratio Literature Review

2.4 Fluid Contact Angle Literature Review

2.5 Dielectrophoresis (DEP) Focusing Literature Review

2.6 Acoustic Focusing Literature Review

2.7 2D Hydrodynamic Focusing Literature Review

2.8 3D Hydrodynamic Focusing

2.9 Optical Flow Focusing Literature Review

2.10 Inertial Flow Focusing Literature Review

Application of Hydrodynamic Focusing in Flow Cytometry and Solve Partial Differential Equations using Machine Learning by Jian Bin Lin A Thesis presented to The University of Guelph In partial fulfilment ofrequirements for the degree of Master of Applied Science in Engineering (Collaborative Specialization in Artificial Intelligence) Guelph, Ontario, Canada © Jian Bin Lin, January, 2024 Abstract Application of Hydrodynamic Focusing in Flow Cytometry and Solve Partial Differential Equations using Machine Learning Jian Bin Lin Advisor: University of Guelph, 2024 Dr. Rafael Santos Microfluidic flow cytometry is gaining popularity as it is portable and reduces the usage of sample and sheath fluid. However, the design of these devices has primarily relied on empirical and trial-and-error approaches, there is a need for a more precise understanding of how fluidic and geochemical parameters affect the device’s performance. To this end, com-computational fluid dynamics simulations were performed herein. When the sheath fluid velocity increases, the width of the focused stream decreases in the square channel before interrogation by the laser beam. modelling results show that the velocity ratio above a certain threshold can achieve the desired final focused stream. The finite Volume Method is typically utilized for solving partial differential equations i.e. momentum equations and energy conservation equations. The problem is that it is time-consuming and computationally expensive. Machine learning has been applied to solve partial differential equations including the Burger equation. Declaration I declare that this written submission represents my ideas in my own words and where others’ideas or words have been included, I have adequately cited and referenced the original sources. I also declare that I have adhered to all principles of academic honesty and integrity and have not misrepresented or fabricated or falsified any idea/ data/ fact/ source in my submission. I understand that any violation of the above will be cause for disciplinary action by the University and can also evoke penal action from the sources which have thus not been properly cited or from whom proper permissions have not been taken when needed. Jian Bin Lin Acknowledgements It is a great pleasure [for me to express my respect and deep sense of gratitude to my M.A.sc](for me to express my respect and deep sense of gratitude to my M.A.sc) supervisor Dr. Rafael Santos, Associate Professor, School of Engineering, University of Guelph, Ontario, for his wisdom, vision, expertise, guidance, enthusiastic involvement and persistent encouragement during the planning and development of this research work. I also gratefully acknowledge his painstaking efforts in thoroughly going through and improving the manuscripts without which this work could not have been completed. I am highly obliged to Mark Bieberich, Mechanical Engineer, Luminex Corporation, help and encouragement for carrying out the research work. I am obliged to my parents Xin Hui Lin and Ya Xiang Yu and my older sister Bi Hong Lin for their moral support, love, encouragement and blessings to complete this task. Jian Bin Lin Table of Contents Abstract…………………………………………………………………………………………ii Declaration……………………………………………………………………………………. iii Acknowledgement…………………………………………………………………………….. iv Table of Contents………………………………………………………………………………. v List of Figures………………………………………………………………………………...viii List ofTables……………………………………………………………………………………x List of Acronyms……………………………………………………………………………….xi List of Glossary……………………………………………………………………………….xiii 1 Introduction…………………………………………………………………………………1 1.1 Fluid Dynamic Focusing Background......................................................................... 3 1.1.1 Governing Equations ...................................................................................... 4 1.2 Machine Learning Background ......

FAQs

What problem does this thesis address in microfluidic flow cytometry design?

It addresses the lack of precise understanding of how fluidic and related parameters affect device performance, since many designs rely on empirical trial-and-error approaches.

How do the simulations explain hydrodynamic focusing behavior?

Computational fluid dynamics simulations show that when sheath-fluid velocity increases, the focused-stream width decreases in the channel before laser interrogation, with the velocity ratio needing to exceed a threshold to reach the desired final stream.

Why does the thesis consider machine learning for solving partial differential equations?

Finite volume methods used for PDEs like momentum and energy conservation are time-consuming and computationally expensive, motivating machine-learning approaches to solve such equations more efficiently.

Aug 5, 2026

Upload by Quinn