Creating nasal cycle simulations by processing MRI and CT scan data with image morphing algorithms

The nasal cycle, characterized by alternating congestion and decongestion of the nasal passages, plays a vital role in nasal function. Predicting the nasal cycle using data from medical imaging modalities, such as magnetic resonance imaging (MRI) and computed tomography (CT), can help elucidate its impact on nasal physiology and inform surgical intervention strategies. This study introduces an image processing algorithm that predicts temporal variations in nasal airway morphology during the nasal cycle by utilizing a single MRI or CT scan from a patient. Our approach pipelines two algorithms: an active contour (snake) algorithm followed by a path planning algorithm. The active contour algorithm identifies corresponding sets of points between contours of the nasal wall and the desired turbinate geometry, while the path planning algorithm generates pathways connecting the corresponding point sets. This process enables the prediction of intermediate geometries between two different levels of nasal congestion observed at distinct time points during the nasal cycle. Prediction accuracy was assessed by comparing predicted and actual intermediate nasal turbinate geometries in scans taken from the same subject at different time points, using a total of six human patients. Two distinct path planning models, linear image morphing and A-star, were evaluated for their accuracy in predicting intermediate nasal geometries at various congestion levels. Cross-sectional area was used to characterize nasal airway geometry. Prediction accuracies for nasal geometries within respiratory regions, including middle and inferior turbinates, ranged from 72.51% - 92.17% for the linear image morphing method and from 70.73% - 90.8% for the A-star method. This algorithm-based tool offers a reliable means to estimate nasal geometries at different congestion levels throughout the nasal cycle using MRI and CT scan data. Coupling this technology with computational modeling could further aid in studying how the nasal cycle influences airflow dynamics under various breathing conditions or pathological states.