Enhancing Analysis of Debris Accumulation in the Post-Lens Tear Film for Scleral Lens Wearers using AI-Driven Quantification
Meng C. Lin, Professor
Optometry
Applications for Spring 2026 are closed for this project.
Scleral lenses, unlike standard contact lenses, are large-diameter rigid lenses that rest on the sclera (white part of the eye) and create a tear-filled reservoir to hydrate the anterior ocular surface. They are primarily recommended for patients with corneal irregularities and dry eye diseases due to their capability to significantly improve vision and comfort. The unique design, featuring a high oxygen-permeable material and a thick fluid reservoir, effectively corrects visual distortions caused by irregular corneas while providing protection and alleviating discomfort for patients suffering from severe dryness. However, a common complication is the accumulation of cellular debris in the fluid reservoir, affecting 26-46% of scleral lens wearers. This buildup can lead to blurry vision, lens fogging, and discomfort, often necessitating frequent lens cleaning and disrupting daily life. Currently, most clinicians rely on subjective grading scales to assess debris severity, lacking standardization. This study aims to develop an AI-based model that analyzes debris characteristics underneath a scleral lens using high-resolution AS-OCT images to directly predict scleral lens-related symptoms. This will help identify debris features that are most strongly associated with patient-reported symptoms and relevant clinical findings.
Role: The undergraduate student will be responsible for the following:
- Understand current study design and study aims
- Assist with developing an AI model to analyze high-resolution AS-OCT images
- Data management
- Help drafting a manuscript for publication upon study completion
This study will be a good opportunity for the student to take part in the process of research planning, data collection, and analysis. The student will gain knowledge in specialty contact lens design, ocular surface physiology, clinical design, and scientific methodology.
Qualifications: Qualifications:
- Background in computer science with experience in AI or machine learning coding
- Strong knowledge of supervised machine learning, particularly classification tasks
- Experience working with image-based data, including image processing and/or computer vision
- Ability to communicate effectively and work both independently and collaboratively within a team
- Curious, proactive, and detail-oriented
- Interest in biology and research is beneficial but not required
Day-to-day supervisor for this project: Bo Tan, Staff Researcher
Hours: 9-11 hrs
Biological & Health Sciences