Selected work / Research archive

Preparing the gallery

Selected work

Research poster index

Open any poster for a high-resolution view. The full-size image loads only when requested.

Research software · Medical imaging · AI

Puti Wen

Puti Wen is a neuroimaging scientist and clinical tool builder at the NYU Abu Dhabi Center for Brain and Health, based in Abu Dhabi, UAE. She builds AI-assisted clinical imaging tools, MRI and neuroimaging pipelines, research automation systems, data infrastructure, and interactive scientific visualizations.

The normal version of this site is an immersive, explorable 3D research gallery. JavaScript is unavailable, so the same professional evidence and ordinary project links are provided below.

Relevant opportunities

Puti is open to relevant opportunities in research software engineering, medical imaging and medical AI, imaging informatics, healthcare data engineering, scientific software, MRI pipeline development, medical image analysis, neuroimaging, scientific visualization, and research automation.

Core technical capabilities

MRI and neuroimaging: fMRIPrep, FreeSurfer, TractoFlow, MRIQC, BIDS, dcm2bids, DICOM, HCP, ANTs, and Nextflow.
Research infrastructure and automation: XNAT, Slurm/HPC, Docker, Singularity, GitHub Actions, Supabase, and the Box API.
Application development and analysis: React Native, Expo, Next.js, TypeScript, Godot, Python, and MATLAB.
MRI research hardware: Siemens scanners, Psychtoolbox, EyeLink eye-tracking, and VPixx/ProPixx.

Professional and technical projects

XNAT Platform and Brain Imaging Biobank

Problem: Large neuroimaging datasets required repeatable processing, validation, backup, and cross-site quality control. Contribution: Puti integrated leading open-source neuroimaging tools as containerized XNAT/HPC pipelines, built two in-house pipelines, automated nightly validation and XNAT-to-Box backups, conducted scanner comparisons, and processed 9,000+ HCP and ADNI subjects for a normative brain dashboard. Methods: XNAT, Slurm/HPC, Docker, Singularity, BIDS, fMRIPrep, TractoFlow, MRIQC, HCP, and dcm2bids. Impact: Puti trained 10+ researchers and resolved 25+ reported data-validation and pipeline issues.
LesionView and Longitudinal Report

Problem: Clinicians need to compare multiple sclerosis lesions across timepoints and verify new or enlarging lesions. Contribution: Puti built the lesion-tracking application and an automated longitudinal report, then migrated the MATLAB prototype to a browser-based React Native and Expo application. Clinical relevance: Features were iterated with Yas Clinic neurologists. Puti is a co-inventor on the LesionView provisional patent application filed in May 2026. View a longitudinal-comparison demonstration.
CvsView

Problem: Clinicians need an efficient way to verify the central vein sign and paramagnetic rim lesions in multiple-sclerosis imaging. Contribution: Puti built a web review tool, identified SWI-FLAIR registration artifacts, relabeled affected data, developed a LightGBM classification workflow, and containerized the tool for internal research evaluation and clinician review, piloted with Harley Street Medical Centre. Methods: Medical image registration, classification, web application development, and Docker. View the CVS/PRL review demonstration.
MyPhysio

Problem: Physiotherapy practices need connected tools for exercise prescription, patient tracking, and administrative work. Contribution: Puti conceived the application, led development with two colleagues, and built admin dashboards, patient portals, exercise prescription, pain diaries, and range-of-motion tracking. Methods: React Native, Expo, Next.js, Supabase, Hostinger, GitHub Actions, and Docker. View an application demonstration.
afqView

Problem: Reviewing tractography for 100+ subjects, 20+ tracts, and 10 iterated versions per subject was being done file by file. Contribution: Puti built an interactive 3D web application with side-by-side comparisons, subject navigation, and color-coded availability indicators. Research relevance: The tool supports scalable white-matter tractography quality control. View the tractography visualization demonstration.
Interactive 3D research and software portfolio

Puti created this full-screen, explorable portfolio to present clinical application demonstrations, MRI and neuroimaging work, research posters, scientific figures, films, games, and artwork as a spatial gallery intended for human review.
Shrimp Tanks

A browser-based shrimp simulation game created by Puti and presented in the portfolio's game room.

Research, publications, and poster work

Single-trial fMRI decoding of 3D motion with stereoscopic and perspective cues — first-author 2025 Journal of Neuroscience article and research poster.
Functional localization of visual motion area FST in humans — first-author 2025 Imaging Neuroscience article and Vision Sciences Society Annual Meeting poster.
Enhanced Prediction of Multiple Sclerosis Disabilities Using Tract-Specific Lesion Load and Advanced Quantitative MRI Metrics — 2025 multiple sclerosis neuroimaging poster.
MP2RAGE Ventricle Distance Profiling Improves Classification of Multiple Sclerosis Subtypes — Organization for Human Brain Mapping Annual Meeting poster, 2025.
CVS-View: AI-powered central vein sign identification — MENACTRIMS 2025 poster.
Center for Brain and Health - Research Institute Day Poster — designed for the Center for Brain and Health at NYU Abu Dhabi, 2025.
MS Research Day — event banner made for the National Multiple Sclerosis Society and NYU Abu Dhabi.
NMSS one pager — stakeholder communication piece made for the National Multiple Sclerosis Society.

The gallery also contains selected figures from Puti's papers, a PhD defense screening, a Center for Brain and Health film, visual designs, an art archive, and an interactive brain room.

Preparing the gallery

Walk through the research.

Research poster index

Move like a first-person game.