An AI-driven image-based feedback system for real-time droplet manipulation
Magnetic Digital Microfluidics (MDM) is a method of manipulating droplets with magnetic particles on a Teflon-coated open surface substrate. By applying and moving a magnetic field, these particles can drag discrete droplets to transport or merge with another droplet. The particles could also work a...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Research |
| Language: | English |
| Published: |
Nanyang Technological University
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/156233 |
| _version_ | 1826111709395288064 |
|---|---|
| author | Tang, Yuxuan |
| author2 | Fei Duan |
| author_facet | Fei Duan Tang, Yuxuan |
| author_sort | Tang, Yuxuan |
| collection | NTU |
| description | Magnetic Digital Microfluidics (MDM) is a method of manipulating droplets with magnetic particles on a Teflon-coated open surface substrate. By applying and moving a magnetic field, these particles can drag discrete droplets to transport or merge with another droplet. The particles could also work as a mixer inside the droplet or be extracted out of the current droplet by a permanent magnet or an electromagnet. After transporting – merging - mixing, particles usually need to be extracted out from the current droplet to merge with the next droplet and then transported again. Basically, a fully functional platform needs to complete at least these four droplet behaviors. However, there exists a gap among MDM platforms, even among Digital Microfluidics (DMF) platforms: current platforms lack an effective means to automatically control the action of magnetic droplets and give feedback of each action's result to the software and the user. So far, the majority of MDM and DMF works are done manually by users. It means the users must see and input the droplet's current and next locations by eyes and manually trigger the actuation of the droplet's actions. The users should also keep a person monitoring the progress of the reaction and judge each part's result by themselves. In this dissertation, the author proposes the first AI-aided image feedback system for real-time manipulation of droplets for MDM, which can continuously trace the coordinates of droplets and particles within 20 ms per frame, and send orders to a UNO board to control two stepper motors and one electromagnet for the actuation of particles' movement. Successfully implemented all basic functions and the remedy of droplets’ magnet disengagement. Due to this system's rapid and accurate real-time control of droplets and particles, the ability of particle extraction, and the easy-to-use graphical user interface. This system can provide a reliable implementation for further droplet-based biomedical applications. |
| first_indexed | 2024-10-01T02:54:57Z |
| format | Thesis-Master by Research |
| id | ntu-10356/156233 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T02:54:57Z |
| publishDate | 2022 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1562332023-04-03T00:16:37Z An AI-driven image-based feedback system for real-time droplet manipulation Tang, Yuxuan Fei Duan School of Mechanical and Aerospace Engineering FeiDuan@ntu.edu.sg Engineering::Bioengineering Engineering::Mechanical engineering Magnetic Digital Microfluidics (MDM) is a method of manipulating droplets with magnetic particles on a Teflon-coated open surface substrate. By applying and moving a magnetic field, these particles can drag discrete droplets to transport or merge with another droplet. The particles could also work as a mixer inside the droplet or be extracted out of the current droplet by a permanent magnet or an electromagnet. After transporting – merging - mixing, particles usually need to be extracted out from the current droplet to merge with the next droplet and then transported again. Basically, a fully functional platform needs to complete at least these four droplet behaviors. However, there exists a gap among MDM platforms, even among Digital Microfluidics (DMF) platforms: current platforms lack an effective means to automatically control the action of magnetic droplets and give feedback of each action's result to the software and the user. So far, the majority of MDM and DMF works are done manually by users. It means the users must see and input the droplet's current and next locations by eyes and manually trigger the actuation of the droplet's actions. The users should also keep a person monitoring the progress of the reaction and judge each part's result by themselves. In this dissertation, the author proposes the first AI-aided image feedback system for real-time manipulation of droplets for MDM, which can continuously trace the coordinates of droplets and particles within 20 ms per frame, and send orders to a UNO board to control two stepper motors and one electromagnet for the actuation of particles' movement. Successfully implemented all basic functions and the remedy of droplets’ magnet disengagement. Due to this system's rapid and accurate real-time control of droplets and particles, the ability of particle extraction, and the easy-to-use graphical user interface. This system can provide a reliable implementation for further droplet-based biomedical applications. Master of Engineering 2022-04-08T13:06:40Z 2022-04-08T13:06:40Z 2021 Thesis-Master by Research Tang, Y. (2021). An AI-driven image-based feedback system for real-time droplet manipulation. Master's thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/156233 https://hdl.handle.net/10356/156233 10.32657/10356/156233 en This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Bioengineering Engineering::Mechanical engineering Tang, Yuxuan An AI-driven image-based feedback system for real-time droplet manipulation |
| title | An AI-driven image-based feedback system for real-time droplet manipulation |
| title_full | An AI-driven image-based feedback system for real-time droplet manipulation |
| title_fullStr | An AI-driven image-based feedback system for real-time droplet manipulation |
| title_full_unstemmed | An AI-driven image-based feedback system for real-time droplet manipulation |
| title_short | An AI-driven image-based feedback system for real-time droplet manipulation |
| title_sort | ai driven image based feedback system for real time droplet manipulation |
| topic | Engineering::Bioengineering Engineering::Mechanical engineering |
| url | https://hdl.handle.net/10356/156233 |
| work_keys_str_mv | AT tangyuxuan anaidrivenimagebasedfeedbacksystemforrealtimedropletmanipulation AT tangyuxuan aidrivenimagebasedfeedbacksystemforrealtimedropletmanipulation |