Multiplanar Reconstructed Computed Tomography Images Improves Depiction and Understanding of the Anatomy of the Frontal Sinus and Recess

Abstract

Aims

The use of multiplanar reconstructed computed tomography (CT) images of frontal recess and sinuses was assessed with regard to depiction and understanding of anatomy and effect on surgical approach.

Materials and Methods

Three otorhinolaryngologists and one radiologist read CT scans of 43 patients referred for routine paranasal sinus scans. Spiral (helical) CT scans were obtained and coronal and parasagittal reconstructions were imaged. Three hundred forty-two readings were analyzed. The scans were assessed in the coronal plane and then in the parasagittal plane. The images were assessed for (i) Bent and Kuhn classification of frontal ethmoidal sinus air cells, (ii) size of frontal sinus ostium (assessed as unsure, normal, small, or large), (iii) use of parasagittal scans regarding additional understanding of the anatomy with particular reference as to how the agger nasi cell and frontal ethmoidal cells were arranged in a three-dimensional space, and (iv) if the parasagittal scan and subsequent three-dimensional picture created altered the surgical approach. The first two criteria were assessed in the coronal plane and then in the parasagittal plane.

Results

There was no statistically significant difference between the Bent and Kuhn classification of frontoethmoidal cells on coronal and reconstructed parasagittal images (t-test; p < 0.05). The parasagittal scans were significantly better than the coronal scans for identifying and assessing the size of the frontal sinus ostium (p > 0.001; chi-square test). Assuming an intraobserver change rate (repeat error) of 10% on CT scan observations, an exact binomial test was performed on S-PLUS, which showed that there was a significant (p < 0.001) proportion of observers who changed their rating after looking at the parasagittal scan. There also was significant improvement in observers’ abilities to identify and classify the size of the frontal ostium as reflected by the number of observers who changed from being unsure on the coronal scans to sure on the parasagittal scans. Observers felt that the parasagittal scans improved their three-dimensional understanding of the anatomy of the frontal recess by 58% on a 10-point Lickert scale. In 55% of these observations, the surgical plan was altered by a mean of 70.2% on a 10-point Lickert scale based on additional information obtained by viewing the parasagittal scans.

Conclusions

The three-dimensional understanding of the frontal recess is improved greatly by using both coronal and parasagittal reconstructed images as compared with coronal images alone. This had important implications on the planning of the surgery in the frontal recess.