Related Experiment Video
Updated: Jun 29, 2025
Modeling and Simulations of Olfactory Drug Delivery with Passive and Active Controls of Nasally Inhaled Pharmaceutical Aerosols
Published on: May 20, 2016
Automated surgery planning for an obstructed nose by combining computational fluid dynamics with reinforcement
Mario Rüttgers1, Moritz Waldmann2, Klaus Vogt3
1Jülich Supercomputing Centre, Forschungszentrum Jülich GmbH, Wilhelm-Johnen-Straße, 52425 Jülich, Germany; Institute of Aerodynamics and Chair of Fluid Mechanics, RWTH Aachen University, Wüllnerstraße 5a, 52062 Aachen, Germany; Jülich Aachen Research Alliance, Center for Simulation and Data Science, 52074 Aachen, Germany.
This study introduces a novel method combining AI and fluid dynamics for nasal surgery planning. It optimizes interventions for better airflow and heat exchange, potentially improving surgical outcomes.
Area of Science:
- Rhinology and Biomedical Engineering
- Computational Fluid Dynamics
- Artificial Intelligence in Medicine
Background:
- Septoplasty and turbinectomy success rates are debated, lacking quantitative data for surgical planning.
- Current nasal surgery planning methods require improvement for better patient outcomes.
- Physics-based planning is desirable for optimizing anti-obstructive nasal surgeries.
Purpose of the Study:
- To develop and validate a novel, accurate method for enhancing nasal surgery planning using physics-based simulation.
- To integrate reinforcement learning with computational fluid dynamics for automated, quantitative surgical planning.
- To evaluate surgical intervention plans based on airflow and heat transfer criteria.
Main Methods:
- Developed an automated pipeline using computed tomography (CT) imaging.
- Combined a reinforcement learning algorithm with large-scale computational fluid dynamics (CFD) simulations.
- Quantitatively evaluated surgical plans based on pressure loss (airflow) and temperature increase (heat transfer).
Main Results:
- The method identified optimal intervention sites for deviated septums and bony spurs.
- Algorithm recommendations varied based on prioritizing airflow versus heat transfer.
- For enlarged turbinates, the algorithm suggested varying degrees of turbinate reduction based on weighted criteria.
Conclusions:
- The proposed physics-based AI method offers a quantitative approach to nasal surgery planning.
- This approach has the potential to improve the success rates of septoplasty and turbinectomy.
- The methodology can be extended to optimize other biomedical flow applications.
