Jennifer Kang-Mieler
Stevens Institute of Technology, NJ·Biomedical Engineering
Two eye diseases cause much preventable vision loss: retinopathy of prematurity (ROP), in which abnormal blood vessels grow in the retina of premature babies, and diabetic retinopathy (DR), in which diabetes damages the retina's small vessels. Both are treatable if caught in time, but timely, consistent diagnosis is hard.
- Fields
Biomedical Engineering
Medicine & Clinical Research
Computer Science
- Topics
retinopathy of prematurity
plus disease
diabetic retinopathy
- Grants
- 2 known grants, through Aug 2028
What they work on
- Optimize and test an animal-to-human (A2H) model that converts animal retinal images into human-like images, using an existing retinal image dataset, to enlarge training data for ROP diagnosis.
- Collect longitudinal retinal images of retinopathy of prematurity from two animal models on three imaging platforms, and hand-annotate them to train deep learning diagnostic models.
- Develop and test CRISP, a cross-platform standardization model that evens out image quality differences between camera systems, and measure how this quality control affects ROP diagnosis.
- Validate whether combining A2H-translated images and CRISP standardization improves deep learning diagnosis of plus disease, the treatment-requiring form of ROP.
- Validate a dynamic tracer kinetic model of fluorescein video-angiography against Evans blue leakage and microsphere blood-flow measurements, and test whether it detects vessel changes before retinopathy in diabetic rodents.
- Track retinal vessel permeability and blood flow over time in healthy people and diabetic patients, and test whether the model detects early DR better than OCT angiography and AOSLO.
More about the research
The lab builds imaging and computing tools that read changes in retinal blood vessels earlier and more reliably, combining animal models, patient imaging and mathematical or deep learning models.
Why it matters
Premature infants at risk of blindness could get faster, more consistent ROP diagnosis, including through telemedicine in underserved regions with few specialists. People with diabetes could benefit from earlier detection of retinal vessel damage, when treatment can prevent most vision loss. Using animal images to train human diagnostic models could also help other clinical fields where patient data and expert labeling are limited.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Ipek Oguz
Vanderbilt University, TN·Biomedical Engineering
Develop custom handheld spectrally encoded coherence tomography and reflectometry (SECTR) hardware, an eye-imaging system combining OCT and surface reflectance, for point-of-care imaging of infant eyes.
Yali Jia
Oregon Health & Science University, OR·Biomedical Engineering
Build a 1-MHz, 120° wide-field, high-resolution OCTA system with GPU real-time processing that captures reliable retinal blood-flow images in a single scan, without stitching images together.
Osamah Saeedi
University of Maryland, Baltimore, MD·Medicine & Clinical Research
Confirm with transgenic mouse models that adaptive optics optical coherence tomography (AO-OCT) signals specifically report retinal ganglion cell bodies and vascular pericytes, validating the imaging against gold-standard histology.
The school
Stevens Institute of Technology: 4 labs on PI Pages, $71M a year on research, 72 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.