Skip to main content

Widefield three-dimensional mosaicking of multiple overlapping volumetric datasets

Primary Investigators:
Yuankai Kenny Tao
Brief Description of Project:
The incidence of diabetic retinopathy (retinal damage due to diabetes) has increased significant in recent years. Early diagnostics and therapeutic guidance is currently limited by the ability to visualize changes in retinal vascular perfusion at the retinal periphery. Optical coherence tomography (OCT) enables noninvasive volumetric imaging of surface and subsurface tissue structures with micron-resolution and has become the “gold standard” for ophthalmic imaging and diagnostics. However, the field-of-view of conventional OCT is limited, and multi-volumetric mosaicking of OCT data is required to access the retinal periphery. Novel image-processing algorithms will be developed to identify corresponding fiducials and perform nonlinear registration on overlapping OCT volumes.

Desired Qualifications:
Required: Proficiency with Matlab and/or ImageJ. Preferred: 1+ semester of signals and systems, and image-processing experience.
Nature of Supervision:
Student(s) will be directly supervised by Primary Investigator and his graduate students.
A Brief Research Plan (period is for 10 weeks):
Week 0-2: Introduction to existing algorithms and data.
Week 2-5: Development of novel registration methods.
Week 5-7: Preliminary testing on existing dataset.
Week 8-9: Validate and quantify algorithm performance on new data.
Week 9-10: Compile data, summarize results, submit abstract for relevant conference.
Number of Open Slots: 1
Contact Information:
Name: Yuankai Tao
Department: Biomedical Engineering