Limitations of acoustic rhinometry determined by a simple model

To determine the accuracy of a commercially available transient signal acoustic rhinometer we constructed simple models from tubing of known dimensions. A series of models was constructed in which the main cylinder was constant whilst the aperture of an included constricted area was varied from 0.07 cm2-0.95 cm2. The area reconstruction for each model was then compared with the true area. Two parameters based on those used in clinical practice of acoustic rhinometry were employed to evaluate the error. The results demonstrate that the measurement of both the volume beyond a constriction and the area of the constriction may be associated with systematic errors. The reconstructed profile of each model is similarly erroneous. As the nose presents a proximal constriction, these findings are of considerable relevance to the clinical use of acoustic rhinometry.