Won June (Kevin) Cho

About Won June (Kevin) Cho

Current (active) research interest is: 1) Developing effective and efficient vision (language) foundation models in digital pathology and researching creative ways to apply them in the clinic. 2) Utilizing state-of-the-art image-to-image translation models (mostly diffusion-like models like Schrödinger/diffusion bridges) for virtual staining of histopathology images for extracting pathophysiology of various diseases. Also interested in integrating multimodal medical data (e.g. Spatial transcriptomics, radiological images) to pathology images.

Experience

1. AI Research Scientist

   DEEPNOID

   Feb 2024 — Present · Seoul, South Korea

   - Developing an in-house vision (language) foundation model for digital pathology for diverse downstream use. Exploring various self-supervised methods and different types of backbone architectures. - Developing several digital whole-slide image-based cancer screening products for clinical deployment. Researching methods to boost Multiple Instance Learning-based classification methods through various means. - Working as a Technical Research Personnel (The technical research personnel system is a form of alternative military service that enables individuals to work in small-to-medium sized companies (research institutes) that are designated by the Military Manpower Administration instead of serving in the military.

2. Graduate Student Researcher

   Johns Hopkins Institute for NanoBioTechnology

   Feb 2024 — Present

   Augmenting my Master's thesis project for even more accurate virtual IHC staining of H&E images (in process for a future publication) by applying new diffusion bridge models, utilizing an extended dataset, and developing new evaluation methods.

3. Graduate Student Researcher at Johns Hopkins Institute for NanoBioTechnology (INBT) - Wirtz Lab

   Johns Hopkins Institute for NanoBioTechnology

   Aug 2022 — Feb 2024 · Baltimore, Maryland, United States

   Researched the following: 1) Thesis Project: Built an image-to-image translation model to perform virtual stain-to-stain conversion between IHC and H&E images of the pancreas to allow accurate islet quantification to determine the pathophysiology of Type 1 diabetes. Workflow includes image registration and generative models such as GANs (pix2pix and its derivatives) and diffusion-like models like Schrödinger Bridges (SB). 2) Side Project: Building and improving a two dimensional workflow for analyzing skin tissue H&E images. Morphometric features found in the images are then extracted to find the cellular biomarkers of aging. Workflow includes image registration and image segmentation models (UNet++, DeepLabV3+, etc).

4. Oncology Data Science Summer Intern at Novartis Institutes for BioMedical Research (NIBR)

   Novartis Institutes for BioMedical Research (NIBR)

   Jun 2022 — Aug 2022 · Cambridge, Massachusetts, United States

   Researched and programmed a specific process in R to analyze/refine gene signatures and calculate gene signature scores from different kinds of cancer patients' bulk RNA-Seq data and pseudobulk single cell RNA-Seq (scRNA) data to deconvolve the application of gene signatures in clinical biomarker analysis.

5. Undergraduate Research Assistant at Johns Hopkins Institute for NanoBioTechnology (INBT)- Mao Lab

   Johns Hopkins Institute for NanoBioTechnology

   May 2021 — Aug 2022 · Baltimore, Maryland, United States

   - Collaborated on projects focused on drug/gene delivery using lipid/polymer nanoparticle-based immunoengineering approaches. -Some major projects are: 1) Development of an Oral Liver-targeted Prime-and-trap Malaria pDNA Vaccine and 2) Development of an Enhanced mRNA Cancer Vaccine via Direct Transfection to Host Antigen Presenting Cells. - In Project 1): • Formulated and screened ~1080 LNPs via cluster-mode in vitro transfection assays and in vivo intrahepatic/intraduodenal injections. • Tested Ai9 Cre reporter mice for co-delivery of anti-inflammatory siRNA with pDNA for in vivo assays of top formulations. - In Project 2): • Formulated and screened ~1080 LNPs via cluster-mode in vitro transfection and immunostimulatory assays on dendritic cells. • Tested Ai9 Cre reporter mice for mRNA delivery and participated in subsequent therapeutic and prophylactic mice tumor studies.

6. Data Science Summer Intern

   Cowell Biodigm Inc.

   Jun 2020 — Sep 2020 · Seoul, South Korea

   Researched several inhibitors that target oncogenes/tumor-suppressor genes and programmed an independent project in R where differential gene analysis of patient RNA-Seq data was followed by pathway analysis to initially screen for targets for new compounds.

Education

• The Governor's Academy

  High School Diploma

  2014 — 2018

• Johns Hopkins Whiting School of Engineering

  Master of Science - MS · Chemical and Biomolecular Engineering

  2022 — 2023

• Johns Hopkins Whiting School of Engineering

  Bachelor of Science - BS · Chemical and Biomolecular Engineering

  2018 — 2022

Skills

• Deep Learning
• Differential Gene Expression (DGE) Analysis
• Cell Culture
• Immunoassays
• Linux
• Nanoparticle Formulation
• MATLAB
• RNAseq Analysis
• Generative Models
• Digital Pathology
• Slurm Workload Manager
• Vision Language Models
• Machine Learning
• PyTorch
• Cancer Immunotherapy
• Data Science
• Python
• R