Gari David Clifford
Emory University, GA·Biomedical Engineering
Serious medical conditions are often caught too late because good screening tests are expensive, hard to reach, or work less well for some groups of people. The two projects described here pair low-cost sensors with machine learning, which are computer methods that learn patterns from data.
- Fields
Biomedical Engineering
Computer Science
Medicine & Clinical Research
- Topics
Alzheimer's disease
preclinical Alzheimer's
digital biomarkers
- Grants
- 2 known grants, through Aug 2027
What they work on
- Test whether scores on VisMET, an eye-tracking memory test, predict global cognition on the Montreal Cognitive Assessment and flag preclinical Alzheimer's disease defined by spinal-fluid biomarkers in older adults.
- Determine whether VisMET predicts cognitive decline over two years and whether age, sex, race or education bias its performance, using Emory participants and the national Bio-Hermes cohort.
- Use machine learning to pull additional cognitive measures from VisMET recordings and test how well they agree with standard neuropsychological tests.
- Validate a machine learning algorithm that detects intrauterine growth restriction from low-cost Doppler ultrasound against standard 2D fetal imaging, in pregnancy cohorts in rural Guatemala and urban Georgia.
- Validate Doppler-based algorithms for maternal hypertension and preeclampsia against gold-standard blood pressure devices and clinical risk tools, then run them in real time on edge-computing hardware for field testing.
More about the research
One project aims to detect early Alzheimer's disease from eye movements. The other aims to screen pregnancies for dangerous conditions using cheap handheld ultrasound devices. Both stress testing the tools in diverse and underserved populations.
Why it matters
Early, accurate Alzheimer's detection that works equally well across races, sexes and education levels could help doctors tell normal brain aging apart from disease that is coming, in both clinics and community settings. Affordable handheld Doppler with built-in artificial intelligence could let front-line health workers in low-resource settings screen pregnancies for preeclampsia, high blood pressure and fetal growth restriction. Such screening may reduce maternal and newborn illness and death.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
David Peters
University of Pittsburgh, PA·Genetics
Develop cell-free DNA methylation analysis of maternal plasma that predicts the clinical trajectory of pregnancies diagnosed with mild or severe preeclampsia.
Rhoda Au
Boston University, MA·Public Health & Epidemiology
Collect digital cognitive metrics (dCog) from a digital recorder, digital pen and smartphone apps in the BU ADRC cohort, and compare people who are positive versus negative for the amyloid-tau-neurodegeneration (A/T/N) signature.
Karen Gibbins
Oregon Health & Science University, OR·Medicine & Clinical Research
Externally validate and update a clinical predictive model of individual placental dysfunction risk in a cohort of 640 pregnant people followed through pregnancy.
The school
Emory University: 70 labs on PI Pages, $1.2B a year on research, 286 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.