Biomedical data is generally multi-modal (imaging, genomics, demography), multi-scale (molecule, cell, tissue, organ, organism), and large-scale (Giga-, Tera-, Peta-byte). As quantitative medicine becomes widely accepted, there is a push to integrate different "views" of a patient into a holistic model for identifying, diagnosing, and treating disease. This presents a number of challenges for organizing and displaying this data: How do we visualize the difference between patients described in 10 dimensions? 100 dimensions? How should we combine 50 genomic features with 100 imaging features? These are not straightforward problems.
In this project, a student will work to develop visualization methods for multi-modal, high-dimensional data. He or she will work with real clinical datasets with the goal of presenting relationships, categories, and important diagnostic information in a way that is clear for the lay-person. The overarching goal of this project is to facilitate "hypothesis generation": by looking at complex data in an intuitive way, we can start to make connections and propose experiments that were previously not obvious.
Disciplines: Biomedical Engineering, High-dimensional Data Analysis, Data Visualization, Data Science
Student Skill-Set Needed: Ability to work with and code web-based visualization technologies (D3.js, HTML5, CSS), MATLAB / Python / R tools, or another data processing and presentation software is necessary; An understanding of biomedical imaging and genomics is crucial for recogn
Compensation: Academic Credit, Volunteer, Work Study
Available: Fall, Spring, Summer
For further information on this opportunity, or to apply, contact:
Faculty Member: Scott Doyle
Title: Assistant Professor
Department: Pathology And Anatomical Sciences
Office: 208 Farber Hall, 3435 Main Street, Buffalo, NY 14214