ZemedicAI: Solar-Powered AI Diagnostic Medical Imaging Booth

ZemedicAI is a solar-powered AI diagnostic ecosystem that enables interpretation of X-ray, ultrasound, and skin imaging, including prenatal care for rural women, even without internet or electricity.

Comment

Inspiration

Over 60% of preventable deaths in Africa stem from late or missed diagnoses, especially in underserved rural regions. At age 12, our founder Emmanuel Ezeji nearly died after an accident due to a delayed, inaccurate diagnosis. That moment sparked a lifelong mission — to ensure no African suffers or dies because of where they live.

ZemedicAI was born to bring hospital-grade diagnostic tools to the last mile — powered by AI, not just electricity. From tuberculosis to pregnancy complications, early detection can save millions of lives.

What it does

ZemedicAI is a solar-powered, modular diagnostic booth that uses AI to interpret medical images in real time, completely offline.

Portable X-Ray AI Detects lung conditions like tuberculosis, pneumonia, cardiomegaly, and early-stage lung cancer.

Ultrasound AI Enables maternal screening, fetal monitoring, abdominal scans, and cardiac assessments — even in clinics without doctors.

Skin Imaging AI Uses a high-resolution dermatoscope to analyze lesions, infections, or early-stage cancer signs.

Optional Doctor Connect If a critical case is detected, the booth lets the patient consult with a live doctor who receives the AI results in real time.

Offline First Booth runs on solar with embedded battery storage and edge AI — no electricity, no internet needed.

How we built it

Integrated NVIDIA Clara, MONAI, and TensorRT for real-time image inference. Trained chest X-ray AI on 600,000+ MIMIC & NIH ChestXray14 plus over 15,000 African scans labeled by Nigerian radiologists. Incorporated Butterfly iQ+ SDK for AI-assisted ultrasound diagnostics. Designed a modular booth with foldable ramp, internal lighting, battery seat compartment, and climate-adaptable materials. Built clinician dashboard for fine-tuning AI with local data and monitoring regional health trends. Architected synchronization layer for safe cloud backup once booth reconnects.

Challenges we ran into

Training AI to handle African patient variations (e.g., genetic diversity, differing X-ray noise patterns). Building robust diagnostics for areas with no reliable internet or power. Designing a medical-grade booth that is affordable, portable, and easy to assemble in conflict zones or mobile clinics. Ensuring AI performs equally well for children, pregnant women, and elderly patients.

Accomplishments that we're proud of

5 pilot booths pre-ordered across 4 Nigerian states. Achieved 94.2% training accuracy and 91.6% validation accuracy on African-specific X-rays. Integrated maternal ultrasound AI, enabling prenatal care in areas without OB/GYNs. Recognized as a Top 3 finalist at Slush'D Lagos 2025. Selected to pitch at international stages and accelerator programs.

What we learned

Maternal care is underdiagnosed in rural Africa — ultrasound AI is a game-changer. AI must be region-aware, not just accurate — involving local clinicians improves trust and performance. Constraints drive innovation. Building offline-first, power-efficient systems for Africa helped us design smarter, lighter models.

What's next for ZemedicAI: Solar-Powered AI Diagnostic Medical Imaging Booth

Scale maternal and fetal care booths to 50+ underserved clinics by Q2 2026. Partner with NGOs and state governments to deploy in IDP camps, refugee zones, and mobile health fleets. Expand diagnostic models for abdominal scans, liver/kidney anomalies, and cardiac abnormalities. Launch Africa’s largest open-access medical image repository to power next-gen AI for global health.

Share this project:

Updates