Last-Mile Vision Care, Powered by AI

Developed by Spectrum Lab, Department of Electrical Engineering, Indian Institute of Science, Bengaluru.

The ChákṣuAI project aims to develop an open-source, AI-powered assistant to support ophthalmologists in diagnosing and managing various eye diseases, including:

• Diabetic Retinopathy
• Glaucoma
• Age-related Macular Degeneration (AMD)

Using state-of-the-art deep learning models, ChákṣuAI provides functionalities such as:

• Classification of retinal diseases (e.g., diabetic retinopathy, glaucoma).
• Detection of abnormalities (e.g., microaneurysms, haemorrhages).
• Retinal Segmentation (e.g., optic disc and cup).
• Volumetric Analysis of fluid in optical coherence tomography (OCT) images.
• Risk Prediction for disease progression.

These capabilities are intended to be integrated into a unified platform to enhance diagnostic accuracy and efficiency, particularly in low-resource settings.

• Project Website: https://chaksu.ai/
• Chákṣu Dataset Paper (Nature Scientific Data): https://www.nature.com/articles/s41597-023-01943-4
• AI for Eyecare Workshop Page: https://chaksu.ai/ai_for_eyecare.html
• Try ChákṣuAI: (Coming Soon...) 🔥

(Note: As of April 2025, these dates are in the past)

• January 20 - 25, 2025: SPARC Course on "Human Perception and Image Analysis in the Era of Machine Learning" held at IISc Bengaluru. Instructors: Prof. Thrasyvoulos Pappas (Northwestern University, USA) and Prof. Gaurav Sharma (University of Rochester, USA). Registration Link.
• January 17, 2025: A One-Day Symposium on Artificial Intelligence for Eyecare held, supported by SPARC, Ministry of Education, Govt. of India and Kotak IISc AI-ML Centre. Registration Link.
• January 14, 2025: The new ChákṣuAI project website was launched!
• November 25, 2024: ChákṣuAI project received funding from the Kotak IISc AI-ML Centre (KIAC) to further develop deep technology solutions for public good.

• Spectrum Lab
• Department of Electrical Engineering
• Indian Institute of Science, Bengaluru
• Project Website: https://chaksu.ai/