Cephalometric x-ray analysis is a fundamental protocol in orthodontics; being annually performed ca. 10 million times; each with ca. 10 minutes effort. Anatomical landmarks must be localized precisely by the specialist and geometries (angles, ratios, distances) are derived to quantify a patient’s maxillo-facial morphology for further diagnosis and therapy planning.
By using massive artificial neural networks (ca. 1/400 of the number of neurons in a human brain), CellmatiQ automated the cephalometric analysis to one second. The training data comprise approx. 3,000 x-rays; analyzed and statistically validated in peer-reviewed campaigns by a large expert group. This “AI-as-a-Service” is medically certified and productive since 2019. The “Arnold Biber Prize” of the German Orthodontic Association awarded for the scientific validation confirmed that the AI “is in no way inferior to the points determined by doctors”.
CellmatiQ’s platform enables systematic data management for AI training and the efficient production of AI analysis modules for use via the Internet or direct integration into medical systems. Further applications under development are e.g. the identification / classification of caries and the pathology detection in panoramic x-rays.
AI has arrived in dentistry and will increasingly support dentists with better process efficiency and a consistently high-quality level.
Written by: Jaroslav Bláha, CEO & Co-Founder, CellmatiQ GmbH
Published by: Network For Innovative Corporate Governance (NICG) – A Research Cooperation between the University of St.Gallen and Swiss Re.
Source of article: https://static1.squarespace.com