Hi, I’m Yoshihiro (Yoshi)
Machine Learning Engineer

Nice to meet you! I am a graduate student using deep learning to denoise and segment images of bone in the Fracture and Fatigue of Skeletal Tissues Lab at UC San Diego. I'm graduating with my PhD in December 2024 and am looking for opportunities to continue my passion for machine learning and engineering.

Contact
Profile Picture

Who am I?

I enjoy coding deep learning and machine learning solutions to grand problems. So far, I have focused on denoising and segmenting micro-CT images to reduce the radiation dose required for each scan. I love collaborating at the crossroads of several fields!

Spending time with family and friends is super important for me. As far as hobbies, I currently love running and playing the game Slay the Spire, though I also actively play tennis, climb, and ski! Going into the outdoors for hiking is the ideal way for me to spend the weekend.

About Me

Education

2023-2024
University of California San Diego
Doctor of Philosophy (PhD), Mechanical and Aerospace Engineering
2020-2022
University of Utah
Graduate Certificate in Deep Learning in AI and Robotics
2019-2022
University of Utah
Master of Science (MS), Mechanical Engineering
2015-2019
North Carolina State University
Bachelor of Science (BS), Aerospace Engineering
Sep 2023 - Dec 2024
Research Assistant, UC San Diego
Fracture and Fatigue of Skeletal Tissues Lab
Jun 2019 - Sep 2023
Research Assistant, University of Utah
Fracture and Fatigue of Skeletal Tissues Lab
May 2018 - Aug 2018
Damage Tolerance Research Intern
NASA Langley Research Center

Experience

Over 6+ years of research experience at multiple universities and NASA, and currently seeking opportunies to contribute to an impactful technology company. I implement and develop machine learning solutions for scientific images now, but am open to application of deep learning in all spaces. So far, I've primarily worked on CT scans, which you can read about in the projects section below. I also have extensive experience performing mechanical tests, such as toughness and fatigue tests at UC San Diego, University of Utah, and NASA. I tested the structural integrity of bone and how it is compromised by disease.

Linkedin Download My Resume

Selected Projects

Deep learning for radiation dose reduction

Radiation dose for CT imaging of bone samples was reduced by 80% using a deep learning approach. This enabled an in situ test that revealed the collagen's important in bone's fracture resistance at the microscale.

Full Study
Deep learning for radiation dose reduction
Self-supervised denoising of scientific images

Self-supervised denoising of scientific images

Self-supervised learning was used to denoise CT images of bone with no high-quality target images required. This method reduced dose imparted by a CT scan by 66%.

Full Study

Data analysis of bone microstructure

Data analysis of thousands of voids and the complex vascular networks in bone using Python (Pandas) revealed the complex relationships between disease and microstructure, with new findings published in several journal articles.

Google Scholar
Data analysis of bone microstructure

Contact

Phone Number

+1 (252) 622-5924

Email

yoshihiro.obata2@gmail.com