A Deep Learning-Powered Platform for Systemic Disease Screening via Dental Radiographs in Underserved Communities

DentalVision AI is an open-access, HIPAA-compliant deep learning platform designed to empower general dentists to perform early-stage screening for critical systemic diseases using standard dental radiographs. By leveraging the existing dental workforce as frontline screeners, the platform addresses a major national public health gap, particularly for the 57 million Americans residing in Dental Health Professional Shortage Areas (HPSAs).

• Official Website: www.dentalvisionai.org
• Research Preprint: medRxiv — MEDRXIV/2026/348653

Millions of Americans suffer from undiagnosed systemic conditions like osteoporosis, diabetes, and cardiovascular disease. Dentists see millions of patients annually for routine care and are uniquely positioned to identify early indicators of these diseases in dental x-rays — but lack the specialized tools to do so.

DentalVision AI analyzes standard panoramic and bitewing radiographs to detect:

• Mandibular Cortical Thinning — indicator of osteoporosis
• Arterial Calcifications — linked to cardiovascular disease risk
• Bone Loss Patterns — associated with uncontrolled diabetes
• Interproximal Caries — with 91% specialist-level accuracy

• AI Models: EfficientNetV2 CNNs trained on 50,000+ anonymized radiographs
• Backend: Python, TensorFlow, Google Cloud Platform
• Compliance: HIPAA-compliant architecture

• Phase 1: Model development and validation
• Phase 2: Research preprint publication (medRxiv)
• Phase 3: Open-source frontend web application
• Phase 4: Pilot programs with dental clinics in HPSAs

Moura, L. O. (2026). A Deep Learning-Powered Platform for Systemic Disease Screening via Dental Radiographs in Underserved Communities. medRxiv. MEDRXIV/2026/348653

CC BY-NC-ND 4.0 — See LICENSE.md for details.

Author: Leonara de Oliveira Moura, DDS | drleonara@dentalvisionai.org