Abstract
The nose performs multiple important functions, including olfaction, humidification and thermal conditioning of the inhaled air and filtration of harmful particles before these reach the lower parts of the air tract. Nasal airway obstruction (NAO) is worldwide one of the most frequent complaints encountered by ear-nose-throat (ENT) physicians. It is a common health condition that affects all age groups and reduces overall quality of life. The functions depend on the airflow through the nasal cavity and will be impacted in case of impaired patency. Within this project the decision support system's proof of concept is further developed to a prototype, and in a second phase into a minimal viable product (MVP). Such a system would allow rhinologists to capture a complete picture of a patient's nasal pathology, e.g. a septum deviation or perforation and enlarged nasal turbinates, which would be in large contrast to current objective measurements, like acoustic rhinometry and rhinomanometry. Unlike the use of audiometry in hearing loss, no gold standard currently exists for assessing nasal function impairment. Clinical examination is mainly used to make treatment decisions but frequently fails to pinpoint the cause of perceived nasal obstruction for a given patient. The system which is envisioned would support physicians (information augmented) while diagnosing and determining the optimal plan of action on patient-specific base, by delivering consistent objective data of nasal airflow and function before surgery (e.g. how the operation will affect other secondary functions of the nose).
Physics-based models, like computational fluid dynamic (CFD) models, have the potential to increase the number of successful operations (i.e. without any recurring or new symptoms in a time span of multiple years). Because of large interpatient variability in nose geometry, a patient-specific approach is required. The manual workflow in creating geometric nasal models remains costly and it requires specific technical expertise that is not available to most physicians. The manual editing is not only tedious but it is also prone to the introduction of errors. The dynamic nature of the nasal mucosa can obscure the true cause of a patient's complaint, making support systems that ignore this and only use single snapshot of the nasal geometry less effective. CFD can also be coupled to different physical laws, such as particle deposition and heating loss, giving a much a broader view on all important nasal physiological parameters.
The final goal of this project is a decision support system that objectively and quantitatively scores nasal (dys)function based on fluid dynamic simulations, while taking the dynamic nature of the internal nose into account. To this end, a multitude of techniques will be used, for example machine learning to automate the extraction of crucial geometric information out of tomographic data and objective geometry characterization to capture the dynamic nature of the patient-specific internal geometry. An important part of the project is the validation of the different innovations using clinical data.
Researcher(s)
Research team(s)
Project type(s)