Analyse der Korrelation von präoperativer Bildgebung mittels Cone Beam CT und intraoperativen Ergebnissen

Digital volume tomography (DVT), also called Cone beam computed tomography (CBCT), is becoming an increasingly popular imaging modality in oral and maxillofacial surgery, as well as ear, nose and throat diseases. CBCT provides a three-dimensional view of the anatomy with lower radiation exposure compared to computed tomography (CT), making it an excellent tool for preoperative planning of middle ear surgeries. However, CBCT is limited in its ability to assess soft tissue due to its focus on high-contrast imaging of bony structures. Nevertheless, the goal of any imaging technique is to present clinical findings as a basis for further clinical decision-making. In this research, we investigated the reliability of preoperative DVT images of the petrous bone, as well as the agreement between radiological findings and intraoperative observations in patients with chronic otitis media. We examined various factors that may influence the assessment of preoperative CBCT images of the petrous bone, including the all-time availability of raw data for reconstruction in different planes (axial, sagittal and coronal). Additionally, we analyzed the influence of various pathologies (otitis media epitympanalis = cholesteatoma vs. chronic otitis media mesotympanalis vs. otitis media adhaesiva) on the assessment of CBCT images. We also investigated the relationship between preoperative CBCT findings and intraoperative anatomy/pathologies. This study includes CBCT image datasets from 85 patients who underwent CBCT imaging as part of preoperative diagnostics before middle ear surgery. Both the image datasets and surgical reports were analyzed with regard to 24 anatomical parameters. The mean age of the patients was 40.4 years and the most common clinical/postoperative diagnosis was cholesteatoma (72.9%), followed by otitis media mesotympanalis (18.8%). Raw data were available for 31.8% of the datasets, allowing for free rotation and reconstruction. Our analysis shows that the structure of the tympanic part of the facial nerve is significantly better assessable with available raw data (p-value 0.0). However, we did not find any significant differences in the ability to assess images in relation to different diagnoses. The auditory ossicles showed the best sensitivity results (with a total value of 0.8-1), except for the incus-stapes joint (0.55) and tympanic part of the facial nerve (0.5), which showed the lowest overall sensitivity. The specificity of the CBCT was highest for the sigmoid sinus, stapes head, semicircular canals, labyrinthine and mastoidal part of the facial nerve (value 1), and lowest for the round (0.32) and oval window (0.49), as well as the posterior limb of the stapes, which in turn means that these were wrongly seen as pathological in more than half of the images. Structures that were seen as intact in the imaging were mostly confirmed as such intraoperatively (negative predictive value), without the completeness of the volume data having any influence. Depending on the diagnoses, the cholesteatoma showed the lowest overall sensitivity, especially in the structures of incus-stapes joint and tympanic part of the facial nerve, as well as the greatest limitation in visualizing the ossicular chain. These limitations may be attributed to the difficulty in distinguishing pathological osteodestructive processes from normal bony anatomy or variants. In summary, it can be confirmed that CBCT is a useful tool for preoperative planning of middle ear surgeries, with some limitations in cases of pronounced osteodestructive inflammation. As technical development of the devices continues to advance, we may be able to further differentiate structures and thus provide even more optimized patient care.