Dermatology Resident · AI Researcher · Writer
Ramie Fathy, MD
Dermatology resident at Johns Hopkins. I build clinical AI tools and study how they should, and should not, be used at the bedside.
› building clinical AI tools. testing whether they belong at the bedside.
› evaluating vision-language models on dermoscopic images…
Software
Browser-based clinical instruments — privacy-preserving, no account required, built around real clinic and resident-education workflows.
Skinoculars
Interactive 3D dermatology visualization tool powered by Three.js and React.
react · three.js · vite02runningAtlasSkin 3D — DEJ Explorer
Exploratory interactive teaching model of the dermal–epidermal junction across tissue, ultrastructural, molecular, and cell scales; not biologically validated or intended for clinical use.
react · three.js · semantic 3d03runningDermatopathology Navigator
Modern React-powered learning platform with AI study assistant, spaced repetition, multiple view modes, and comprehensive differential diagnosis builder.
react 18 · gemini ai · webgl shadersSelected publications
Comparative Assessment of GPT-4 Vision in Dermoscopic Image Analysis
journal of the american academy of dermatology2022Pediatric teledermatology: a retrospective review of 1199 encounters during the COVID-19 pandemic
journal of the american academy of dermatology2022Lack of skin of color in Google image searches may reflect under-representation in educational resources
journal of the american academy of dermatology2020Climate change and inpatient dermatology: A call to action
current dermatology reportsAbout
I am completing my final year of dermatology residency at Johns Hopkins, with clinical interests in inflammatory and complex medical dermatology, dermatopathology, and clinical AI evaluation.