Our work on comparing concepts used by CNNs to detect glaucoma with regions fixated most by experts (via measuring their eye movements) on OCT images appeared in IEEE TBME earlier this month! :) ieeexplore.ieee.org/document/92864…

Our latest research on "A multimodal deep learning system to distinguish late stages of AMD and to compare expert vs. AI ocular biomarkers" has just been published with Springer Nature in @ScientificReports!

Check it out here: nature.com/articles/s4159…

🎓👏Congratulations to Columbia BME's Kaveri Thakoor on a successful doctoral dissertation defense: Robust, Interpretable, and Portable Deep Learning Systems for Detection of Ophthalmic Diseases (PIs: Paul Sajda, PhD LIINC & Donald Hood, PhD). Well deserved, Kavi!

Kaveri Thakoor on #AI for eye disease detection #glaucoma #AMD #datascienceday2023

Celebrating the recent completion of our Research Topic on the 'Role of Eye Movements in Vision, Attention, Decision-making, and Disease' successfully led by Profs. Kaveri Thakoor, Fatema Ghasia and Hantao Liu. 👁️Explore the collection here: fro.ntiers.in/RT55844

Big news: Kaveri Thakoor received the prestigious RPB Tom Wertheimer Career Development Award in Data Science to support eye research! This award is a testament to the impact her work will have on people around the world.

Kaveri Thakoor created an AI-powered tool to help in diagnosing eye disease. It's designed to act as a partner—trained by physicians for physicians—& is capable of explaining its diagnostic choices, akin to a human clinician Read about her important work: datascience.columbia.edu/news/2024/a-da…

🚨 New publication alert: 📢 “AI-CNet3D: An Anatomically-Informed Cross-Attention Network with Multi-Task Consistency Fine-tuning for 3D Glaucoma Classification.” 🖊️ R Kenia, A Li, R Srivastava, Kaveri Thakoor. ⬇️

🎯 Authors propose a hybrid 3D CNN with cross-attention for glaucoma classification from OCT scans, designed to capture asymmetries across hemiretinas and integrate optic nerve head and macula features. 🔎 Free to read: doi.org/10.59275/j.mel…