VoxelMD
Autonomous Radiology Intelligence
Clinical-grade AI that reads medical images and drafts diagnostic reports. Founded by a practicing Neuroradiologist.
The Problem
Radiology is in crisis. The world is running out of radiologists.
The Solution
Two complementary products that automate radiology — today and tomorrow.
Voxel Suite
Workflow automation that eliminates 80% of radiology dictation time. Auto-injects clinical history, generates prior imaging summaries, and pre-populates report templates.
Voxel Vision
Autonomous AI that reads medical images and drafts preliminary diagnostic reports. Currently prototyped for Thyroid Ultrasound and CT Perfusion studies.
Traction
Real users, real metrics, real clinical validation.
Market Opportunity
Global diagnostic imaging AI market by 2030 (Allied Market Research).
North American radiology reporting automation — CT, MRI, Ultrasound workflows.
US private radiology groups automating Thyroid US, CT Perfusion, and general reporting.
Revenue Model
B2B SaaS
Per-radiologist subscription for Voxel Suite workflow automation. Targeting $200–500/mo/seat for private radiology groups and hospital systems.
CPT Code Billing
Once Voxel Vision achieves FDA clearance, each AI interpretation generates a billable CPT code — aligning revenue directly with imaging volume. Highly scalable with near-zero marginal cost.
Technology
Two deployment pathways — customers choose the architecture that fits their security and budget requirements.
On-Premise (NVIDIA GPU)
Open-source VLMs (Gemma 4 / MedGemma, Qwen 3.5 VL) run locally on hospital-owned NVIDIA GPUs. Zero PHI leaves premises. Air-gapped architecture.
HIPAA Cloud (BAA)
De-identified DICOM data sent to BAA-covered Vertex AI (Gemini 3 Flash). Same DICOM-in/results-out pattern used by RapidAI and Rad AI. Lower cost, no GPU hardware required.
HIPAA Compliant
Both pathways meet HIPAA requirements
Epic + PACS
Deep EHR & imaging system integration
Customer Choice
Buy GPU hardware or use managed cloud
Founder
Mo Fakhri, MD
Founder & CEO · Practicing Neuroradiologist · Full-Stack Developer
• Harvard Medical School (postdoctoral research), UCSF (Neuroradiology fellowship), Mallinckrodt Institute of Radiology (residency)
• NIH T32 Research Grant and RSNA Research Resident Grant recipient
• Writes every line of production code — Python, TypeScript, CUDA, model training pipelines
• Reads thousands of studies annually as a practicing radiologist — builds AI from real clinical insight, not synthetic datasets
The Ask
Seeking pre-seed investment to accelerate product development, expand the team, and initiate the FDA pre-submission pathway.
Hire ML engineer and clinical validation lead
Scale Voxel Vision to 5+ imaging modalities
Pre-submission meetings and 510(k) preparation
Let's Talk
Interested in VoxelMD? We'd love to connect.