Deep Reasoning AI Virtual Anatomical Technology & Augmented Reasoning Platform: Application to Craniomaxillofacial Reconstruction

Deep Reasoning is developing the Deep Reasoning AI Virtual Anatomical Technology & Augmented Reasoning (DR-AIVATAR) platform, a multi-agent AI system for high-precision personalized 3D anatomical modeling, planning, reconstruction, and decision-making. DR-AIVATAR will provide hospital and clinic customers with accurate, rapid, low-cost, personalized 3D anatomical modeling, surgical planning, and decision-making capabilities. The initial focus will be on high-precision, personalized 3D anatomical modeling and virtual surgical planning (VSP) tools for congenital, acquired, and post-traumatic reconstructions and oncologic resections involving structures of the skull, face and jaw. The midface and mandible are critical components of the facial skeleton, and pathologies affecting these areas can significantly impact a person's ability to eat, provide an adequate airway, perform activities of daily living, appearance, and maintain self-esteem. The current standard of care for treating these pathologies often involves reconstruction with VSP and 3D printing, but access to these technologies is limited to large tertiary centers. Further, challenges with costs, labor, time, and accuracy highlight the need for a more simplified and scalable solution. DR-AIVATAR will encompass AI-driven segmentation algorithms that are accurate for each disease state and will significantly reduce the time to create a 3D model. DR-AIVATAR will aid in more accurate VSP for a diverse population, further reducing the time to develop an operative plan in a timely manner. In addition, DR-AIVATAR will include AI-driven tools to generate VSP plans based on known variables, with minimal human input required, which will allow this planning tool to be used by less experienced users and decrease the time experienced users engage in planning, while also improving the quality of the procedures and outcomes. All of these characteristics will significantly reduce the learning curve to perform VSP, reduce staff requirements, reduce training time for new employees, and finally, but most importantly, improve the lives of patients afflicted with these disorders by democratizing the technique to those with less training no matter where they are on located. For this Phase I project, Deep Reasoning will 1) Develop and demonstrate the ability of DR-AIVATAR to perform AI-driven 3D modeling for craniomaxillofacial CT scans (below the mandible to the top of the head) and lower extremity CT angiogram runoff, 2) Demonstrate the generalizability of DR-AIVATR through the development of age, gender, height, and racially specific normative skull models combining AI and statistical shape modeling, and 3) Use DR-AIVATAR to develop AI-Assisted Virtual Surgical Planning tools for congenital, acquired, and post-traumatic craniomaxillofacial reconstructions and oncologic resections involving the mandible and midface including fibula free flap reconstruction. Project Number: 1R41DE034646-01 | Fiscal Year: 2025 | NIH Institute/Center: National Institute of Dental and Craniofacial Research (NIDCR) | Principal Investigator: Jonathan Morris (+1 co-PI) | Institution: DEEP REASONING AI INC, SARASOTA, FL | Award Amount: $350,000 | Activity Code: R41 | Study Section: Special Emphasis Panel[ZRG1 ISB-Z (10)] View on NIH RePORTER: https://reporter.nih.gov/project-details/11069219