Summary
Last updated: 2024-02-01
A student at the Stanford University, undertaking a mechanical engineering PhD course.
Publications
Machine learning-aided three-dimensional morphological quantification of angiogenic vasculature in multiculture microfluidic platform
Experience
• Developed a deep learning network for the virtual staining of nuclei in on-chip angiogenesis, based on a GAN model with advanced UNet generator. • Developed a cell tracking network based on the Graph Laplacian to track nuclei of the virtually stained angiogenesis, enabling dynamic evaluation of angiogenesis.
• Treated microplastics on a vasculogenesis microfluidic platform, quantifying the microvasculature's response to microplastics through fluorescent staining and RT-qPCR.
• Developed a graph convolutional network consisting of edge convolution and cascaded attention module and improved the deep learning network’s skeleton segmentation capacity. • Proposed and implemented a point cloud base 3D analysis pipeline optimized for quantifying angiogenic vasculature in MV‑IMPACT platform and achieved a 47.9% reduction of error over the conventional maximum intensity projection analysis method on average.
• Investigated the effects of microplastics on the vasculature within the chip by introducing microplastics of different concentrations and sizes to the on-chip microvasculature. • Quantified the response of the on-chip microvasculature to microplastic treatment using immunostaining and RT-qPCR.
• Designed a microfluidic device that leverages spontaneous capillary flow under hydrophilic conditions through rapid prototyping, allowing for selective patterning of hydrogels in specified regions and co‑culture of two or more cell types. • Demonstrated the effects of bacterial stimulation on tumor spheroid and corresponding pro‑inflammatory response of macrophages experimentally, and therefore emulated the fundamental constituents of bacteria‑colonized tumor‑microenvironment in vitro.
• Established a novel protocol to stabilize photopolymerized poly (ethylene) glycol diacrylate (PEGDA) microfluidic device for cell culture. • Constructed a computationally automated diffusion switch system by controlling fluid inflow using a syringe pump and designed a low‑pass filter system that can selectively filter lightweight molecules based on their diffusion coefficient.
Honors and Awards
Bucheon Jang-hak Foundation
Two semesters, 50% of tuition
Seoul National University
30% of tuition
Seoul National University
50% of tuition
Seoul National University
50% of tuition
Seoul National University
Full-tuition
Seoul National University
• Led a team of six and developed the ball classifier machine that can assort balls based on their weight, up to three different types.
• Took 1st place among 16 teams composed of 112 students.
Sangjin Jang-hak Foundation
50% of tuition
Stanford University
$12,800
Stanford University
Full tuition (1 yr)
Extracurricular Activity
• Conducted case law analysis and discussed current affairs in legal interpretation related to it. • Led science and technology-related sessions.
• Managed and advised modeling for 3D printing. • Guided lab tour and explained fundamentals of different 3D printing methods and their application on research.
• Mentored high school students in a one‑on‑one relationship with a monthly conversation on topics in science and mechanical engineering.
• Analyzed and interpreted the collected signal intelligence and reported vital information to the higher command. • Excellence award in military occupational specialty education.
Skills & Proficiency
Language
Python, MATLAB, Verilog, C/C++
Framework
PyTorch, Tensorflow, OpenCV, Open3D, Pandas
3D CAD and Printing Tools
SolidWorks, AutoCAD
Computational Simulation Tools
COMSOL Multiphysics, Acusolve
Bio Experiment
Cell culture, Bacteria culture, Confocal microscopy, ELISA, RT-qPCR
Microfluidic Device Fabrication
PEGDA Photopolymerization, 3D Printing, Laser cutting & engraving
Open Source Projects
A computational tool for 3D analysis of vascular networks
