teaching
Teaching, mentorship, and courses I am prepared to contribute.
Teaching students to connect theory, physics, computation, and judgment
My teaching is guided by a simple principle: students should leave a course able not only to apply methods, but also to understand where those methods come from, what assumptions they encode, and when they should or should not be trusted. In imaging and machine learning, this means connecting mathematical principles, physical measurement, algorithm design, and empirical evaluation rather than teaching each in isolation.
I am especially motivated by courses that help students move between foundations and practice. In medical imaging and signal processing, students often learn to run sophisticated pipelines before they have fully developed the habits of mind needed to interpret outputs critically. I therefore emphasize questions such as: What is the forward model? What information is missing? What assumptions make this method work? How would we know when it fails?
In my teaching, I aim to make these questions part of the normal rhythm of learning. I design assignments and discussions that ask students not only to implement methods, but also to identify modeling assumptions, diagnose likely failure modes, and justify evaluation choices. My goal is to help students develop scientific judgment alongside technical fluency.
I also aim to create an inclusive classroom by providing multiple entry points to technical material, encouraging collaborative problem solving, and making expectations clear and transparent. Where feasible, I use practices that reduce bias in evaluation and support steady feedback, so that students can improve through the course rather than simply being judged at its end.
Recent teaching roles
Head Teaching Assistant, Medical Image Analysis
Johns Hopkins University, Spring 2025
Head Teaching Assistant, Medical Imaging Systems
Johns Hopkins University, Fall 2023
Head Teaching Assistant, Music Signal Processing
Johns Hopkins University, Fall 2022
How I mentor
I mentor students to pair technical depth with research judgment: identify the core scientific question, choose methods that respect problem structure, and evaluate results rigorously. In practice, I set clear milestones, give frequent written and verbal feedback, and use regular check-ins to unblock progress. I also support students in developing as researchers by helping them frame results, prepare talks, and write with clarity.
Courses I would be excited to teach or co-develop
Medical image analysis
Segmentation, registration, reconstruction, harmonization, generative methods, and evaluation on clinically realistic datasets, with emphasis on data fidelity, robustness, and failure modes.
Computational imaging and inverse problems
Foundations of reconstructing images from indirect or incomplete measurements, including forward models, optimization, regularization, and data-consistent learning.
Machine learning for scientific and medical data
Reliability, representation learning, generative modeling, and domain-specific constraints for safety-critical and scientifically grounded applications.