cv

Download .pdf on the right.

Basics

Name Númi Sveinsson Cepero
Label Ph.D. in Mechanical Engineering | Computational Modeling & Medical Imaging
Email numi@berkeley.edu
Url https://numisveinsson.com
Summary Computational modeling and medical imaging researcher specializing in biofluid dynamics and deep learning for cardiovascular applications. Experienced in developing machine learning frameworks for vessel segmentation and automatic vascular model construction. Ph.D. in Mechanical Engineering from the University of California, Berkeley (2025).

Education

  • 2020 - 2023

    Berkeley, CA, USA

    M.S.
    University of California, Berkeley
    Mechanical Engineering
  • 2020 - 2025

    Berkeley, CA, USA

    Ph.D.
    University of California, Berkeley
    Mechanical Engineering
    • Biomechanics
    • Machine Learning
    • Scientific Computing
    • Biofluid Dynamics
    • Finite Element Method
    • Deep Learning for Image Analysis
  • 2019 - 2020

    Boston, MA, USA

    Doctoral Transfer
    Boston University
    Biomedical Engineering (Ph.D. coursework)
  • 2018 - 2018

    Stanford, CA, USA

    Visiting Scholar
    Stanford University
    International Honors Program
  • 2016 - 2019

    Reykjavík, Iceland

    B.S.
    University of Iceland
    Mechanical Engineering

Research

  • 2025 - Present

    Austin, TX, USA

    Postdoctoral Position
    The University of Texas at Austin — Oden Institute for Computational Engineering and Sciences
    Research focuses on integrating artificial intelligence, medical imaging, and computational modeling to study physiology and advance medicine. Expanding research to explore clinically driven applications of AI and simulation, with the goal of developing digital tools that bridge computational models and patient care.
    • Advisor: Prof. Charles A. Taylor
  • 2020 - 2020

    Reykjavík, Iceland

    Research Internship
    Kerecis — Research and Development
    • Advisor: Prof. Sigurður Brynjólfsson
  • 2020 - 2025

    Berkeley, CA, USA

    Graduate Research
    University of California, Berkeley — Mechanical Engineering Department
    Earned Ph.D. in Mechanical Engineering. Work with Prof. Shawn Shadden combined biomechanics, machine learning, and image-based modeling to improve vascular segmentation and simulation. Doctoral projects introduced open-source tools for automated vascular reconstruction and simulation-ready mesh generation from medical images.
    • Advisor: Prof. Shawn C. Shadden
    • SeqSeg: Open-source tool for automated vascular reconstruction
    • MeshGrow: Open-source tool for simulation-ready mesh generation from medical images
  • 2019 - 2019

    Reykjavík, Iceland

    Undergraduate Research
    University of Iceland — Department of Medicine
    • Advisor: Prof. Thórarinn Guðjónsson

Work

  • 2020 - 2020

    Reykjavík, Iceland

    Research Intern
    Kerecis
    Worked on material property analysis for medical biomaterials.
    • Collaborated on R&D for regenerative medical materials.
  • 2020 - 2025

    Berkeley, CA, USA

    Graduate Researcher
    University of California, Berkeley — Mechanical Engineering Department
    Developed deep learning–based methods for automatic vascular model construction and trajectory-oriented vessel tracking in medical imaging. Integrated physics-based modeling and machine learning to enable patient-specific cardiovascular simulations.
    • Developed SeqSeg: a deep learning framework for coronary artery segmentation
    • Created MeshGrow: an automated pipeline for unified cardiac and vascular mesh generation
    • Contributed to SimVascular open-source cardiovascular modeling software
    • Mentored undergraduate and MEng teams in medical image analysis projects
  • 2019 - 2019

    Reykjavík, Iceland

    Undergraduate Researcher
    University of Iceland — Department of Medicine
    Developed computational models for new lung experimental technologies.

Publications

Service

Awards

Skills

Programming and Scripting
Python
C++
MATLAB
Bash
Medical Image Processing
Segmentation
Registration
CT/MRI
ITK
VTK
MONAI
Deep Learning and Computer Vision
CNN
Transformers
U-Net
Gaussian heatmaps
Trajectory prediction
Transfer learning
Computational Modeling and Simulation
Mesh generation
CFD
FSI
SimVascular
ANSYS
ParaView
Software Development and Collaboration
Git
GitHub
Docker
Linux
Continuous Integration
Visualization and Data Analysis
Matplotlib
ParaView
VTK
General Tools and Workflow
LaTeX
Overleaf
HPC
Slurm

Languages

English
Fluent
Icelandic
Native
Danish
Intermediate
Spanish
Intermediate
French
Beginner

Projects

  • SeqSeg
    Deep learning–based framework for automated segmentation of coronary arteries from CT images.
  • MeshGrow
    Framework for unified cardiac and vascular mesh construction from medical images.
  • SimVascular
    Open-source cardiovascular modeling software. Contributed features and documentation.
  • MIROS
    Toolkit linking patient-specific medical imaging data to reduced-order hemodynamic simulations.
  • Vascular Model Repository
    Online platform for sharing and visualizing vascular geometries and simulation data.