Bite-Block Relocation Error in Image-Guided Otologic Surgery

Abstract

Abstract. Otologic surgery is undertaken to treat ailments of the ear including persistent infections, hearing loss, vertigo, and cancer. Typically performed on healthy patients in outpatient facilities, the application of image-guided surgery has been limited because accurate (<1mm), non-invasive fiducial systems for otologic surgery are not available. We have developed such a system, which repeatably attaches to a subject via a dental bite block [1]. In a previous report, ex-vivo validation indicated that mean target registration error ± standard deviation for 234 targets within the surgical field was 0.73±0.23mm with a root mean square of 0.77mm [2,3]. In an effort to further understand the limitations of our system, we undertook, and report herein, a detailed error analysis to determine what portion of the error is attributable to the inaccuracies in relocation fit of the fiducial frame to a subject's dentition. Our analysis shows that the root-mean-square error due to relocation alone is approximately 0.66 mm, while that due to fiducial registration is approximately 0.30 mm.

Clinical determination of target registration error of an image-guided otologic surgical system using patients with bone-anchored hearing aids

“…The error component due to FLE does not change much though because there are no changes expected in the anatomy of the inner ear, and the localization accuracy of the markers remains approximately the same throughout. We also note that the main component of the total error is the bite-block relocation error, as noted for skulls in [6,7].…”

“…The error in reattaching the bite block at the same position relative to the teeth results in error in placing the fiducial frame (virtually) at the same pose (position and orientation) relative to the patient. This biteblock relocation error is a major component of the total TRE measured at any target point [6,7]. This error component will be referred as TRE frame , and can be assumed to be uncorrelated to TRE FLE .…”

“…Clinically, it is important to document the target registration error (TRE) of this system at the surgical target. Skull-based studies conducted previously showed that this system has submillimetric accuracy [3,4,6,7]. The mean TRE in the surgical field was found to be 0.73mm±0.23mm with a root-mean-square value of 0.77mm by implanting markers in cadaveric skulls [3,4,6].…”

“…Skull-based studies conducted previously showed that this system has submillimetric accuracy [3,4,6,7]. The mean TRE in the surgical field was found to be 0.73mm±0.23mm with a root-mean-square value of 0.77mm by implanting markers in cadaveric skulls [3,4,6]. A unique method was described to perform clinical validation of the system without the need for intra-operative data collection, and proof of concept was demonstrated on cadaveric specimens for determining targeting accuracy at the cochlea [7].…”

Quantification and visualization of alveolar bone resorption from 3D dental CT images

“…Befu et al [3] proposed a method to obtain clear threedimensional images from noisy dental CT images. Also, head CT is being used in dental implantology and oral and cranio-maxillofacial surgery [4][5][6][7][8][9][10]. Concerning alveolar bone resorption, Li et al [11] detect and evaluate root decay and bone loss (resorption) on 2D dental X-ray images by a level set method and an SVM (support vector machine) classifier.…”