Research

Broadly, my research interests include applying strong mathematical theory to machine learning contexts to improve results and better understand models. My work focuses on numerical optimization and machine learning with applications in medical imaging. I also have an interest in educational research, using machine learning techniques to better understand student learning in introductory computer science.


My doctoral work focuses on developing an efficient and effective tool for generating accurate and high-resolution human brain images from EPI MRI scans. This work also uses tools and techniques from image registration, optimal transport, and GPU computing to accomplish these goals.


During my undergraduate studies, I engaged in research on a project called Region Radio, an artificially intelligent place-based roadtrip storyteller. I also did some work on a Medical School Education research project looking at using digital flashcards to provide real-time learning feedback to students and professors.

Conference Presentations

• Fast and Trainable Susceptibility Artifact Correction for EPI-MRI. SIAM Imaging Science March 2022.
• Towards Fast Bi-Level Optimization for EPI-MRI Susceptibility Artifact Correction and DTI. SIAM Mathematics of Data Science September 2022.
• epic4D: Efficient Echo-Planar MRI Distortion Correction for Four-Dimensional Diffusion Tensor Images. AMS Sectional Meeting March 2023.
• Efficient Echo-Planar MRI Distortion Correction for Three-Dimensional and Four-Dimensional Images. SIAM Optimization May 2023.

  Conference Proceedings

• Fisher, D., Markert, E., Roberts, A., & Varma, K. (2019). Region Radio: An AI that Finds and Tells Stories about Places. Proceedings of the 10th International Conference on Computational Creativity (pp. 336-340). (link)

  Posters

• Using Anki to Provide Real-Time Learning Feedback in First-Year Anatomy - AMA ChangeMedEd 2019
• Region Radio: An AI that Finds and Tells Stories about Places - 2018 Doctoral Consortium on Computational Sustainability