From Ultrasound to Diagnosis in 48 Hours
MERIT AI is our flagship research initiative — a clinically deployed ophthalmic AI system built to screen, classify, and triage 250+ blinding conditions using a portable ultrasound device guided by on-premise AI.
A Crisis of Access
India faces a critical shortage of ophthalmologists in rural areas, where the 48-hour diagnostic window determines whether a patient retains their sight.
With one ophthalmologist per 70,000 rural residents, delayed diagnosis is the primary cause of preventable blindness across India. The 48-hour window — the critical period where intervention prevents permanent sight loss — is routinely missed. MERIT AI was conceived to bring expert-level diagnostic capability to the point of care, wherever the patient is.
MERIT AI — On-Premise, Zero Cloud
A portable Butterfly iQ3 ultrasound probe feeds into an on-premise AI engine that delivers diagnosis without any data leaving the facility.
- Butterfly iQ3 portable ultrasound — wireless, hand-held
- On-premise AI engine — GPU-accelerated, no internet required
- 4-stage diagnostic pipeline: classify → segment → detect → diagnose
- Zero-cloud, HIPAA & DPDPA compliant
- Teleguidance-capable for remote specialist review
The AIID Research Framework
Application → Implementation → Integration → Dissemination. The four-stage translational philosophy that drives every decision in MERIT AI's development.
Application
25 years of clinical vision translating into a defined problem: preventable blindness caused by delayed ophthalmic diagnosis. The clinical need drives everything.
Identifying 250+ blinding conditions with a 48-hour critical window across under-served regions.
Implementation
Research, phantom lab training, animal testing, and data collection. Phase 1 (Classification) complete at 91% accuracy; Phase 2 (Segmentation) active.
175 phantom images, 600 frames/scan, 10,000-image collection target, De Cure patient pipeline.
Integration
Software development, on-premise deployment, and clinical workflow embedding. Zero-cloud, HIPAA/DPDPA compliant. GPU-accelerated inference on Butterfly iQ3 input.
Docker-orchestrated stack, SHA-256 data security, RBAC, teleguidance-capable interface.
Dissemination
Making the technology available — in rural clinics, District Hospitals, and globally via teleguidance. NIH SBIR grant in pursuit for Phase I expansion.
Athreya Inc. (US) + Validus Institute co-applicant. Target: accessible AI diagnostics worldwide.
Peer-Reviewed Research
Our team's contributions to the scientific literature in medical imaging, AI diagnostics, and ophthalmic research.
2024
Ocular Ultrasonography in the Assessment of Optic Nerve Sheath Diameter: A Systematic Review
Khazaei HM, Abbas K, Oteibi M
Journal of Ophthalmology, 2024
Point-of-Care Ultrasound for Orbital Disease Detection: Current Evidence and Future Directions
Khazaei HM, Munaver J, Abbas K
American Journal of Ophthalmology, 2024
Thyroid Eye Disease: Ultrasonographic Features and AI-Assisted Classification
Khazaei HM, Oteibi M
Thyroid, 2024
2023
Tear Proteomics in Ocular Surface Disease: Biomarker Discovery Using Mass Spectrometry
Khazaei HM
Investigative Ophthalmology & Visual Science, 2023
Non-Invasive Optic Nerve Sheath Diameter Measurement: Validation Against CT in Emergency Settings
Khazaei HM, Abbas K
Emergency Medicine Journal, 2023
Ocular Oncology and Uveal Melanoma: Ultrasound Features and Differential Diagnosis
Khazaei HM
Ocular Oncology and Pathology, 2023
Bioinformatics Approaches to Ophthalmic Disease Classification: A Scoping Review
Khazaei HM, Oteibi M
Frontiers in Genetics, 2023
2022
Retinal Detachment Screening Using Portable Ultrasound in Low-Resource Settings: A Feasibility Study
Khazaei HM, Munaver J, Abbas K
British Journal of Ophthalmology, 2022
Interested in Collaborating?
We welcome research partnerships, clinical data contributions, and institutional collaborations that advance MERIT AI toward full clinical deployment.
Contact Us