Numi Sveinsson Cepero

Hey there! I’m a PhD candidate delving into the exciting world of biomechanical engineering at UC Berkeley. My research journey involves a passionate mix of biomechanics, machine learning, and computer modeling.

Currently, I’m at the crossroads of cardiovascular health, computational modeling, and deep neural networks. I’m teaming up with the Cardiovascular Biomechanics Computation Lab at Stanford University to craft SimVascular, an open-source software for patient-specific blood flow modeling. We’re also building a free, open-access Vascular Model Repository because sharing is caring, especially in the realm of cutting-edge research.

I thrive on collaboration and adore applying engineering and machine learning to real-world problems. Let’s connect and brainstorm some groundbreaking ideas!

Outside the lab, I find joy in exploring history, jet-setting to new destinations, nurturing green life, and whipping up culinary wonders with friends. Life’s an adventure, and I’m here for it all!

news

May 28, 2025	I had the honor of delivering the commencement speech at the UC Berkeley Engineering PhD ceremony on May 24, 2025. It was a wonderful experience to share my journey and insights with the graduates, their families, and friends. The full speech is available on YouTube: Watch the Commencement Speech.
May 2, 2025	Commencement 2025 - PhD Graduation
Jan 7, 2025	I am excited to announce that I have been awarded a 2025 Spring M.E. Department Scholarship by the Department of Mechanical Engineering at UC Berkeley. This scholarship will support my research on the development of computational tools for the analysis of blood flow and medical images. I am grateful to the department for this recognition and support.
Dec 25, 2024	End of FEM Course TA
Nov 22, 2024	Gave lecture in FEM Course

latest posts

May 1, 2025	SeqSeg, Now a Pip Installable Package
Jan 15, 2025	SeqSeg, From New Dataset to Training to Inference
Sep 24, 2024	ParaView Tutorial

selected publications

SeqSeg: Learning Local Segments for Automatic Vascular Model Construction

Numi Sveinsson Cepero, and Shawn C. Shadden

Annals of Biomedical Engineering, Sep 2024

Abs PDF

Computational modeling of cardiovascular function has become a critical part of diagnosing, treating and understanding cardiovascular disease. Most strategies involve constructing anatomically accurate computer models of cardiovascular structures, which is a multistep, time-consuming process. To improve the model generation process, we herein present SeqSeg (sequential segmentation): a novel deep learning-based automatic tracing and segmentation algorithm for constructing image-based vascular models. SeqSeg leverages local U-Net-based inference to sequentially segment vascular structures from medical image volumes. We tested SeqSeg on CT and MR images of aortic and aortofemoral models and compared the predictions to those of benchmark 2D and 3D global nnU-Net models, which have previously shown excellent accuracy for medical image segmentation. We demonstrate that SeqSeg is able to segment more complete vasculature and is able to generalize to vascular structures not annotated in the training data.