Automatic segmentation of intracochlear anatomy in conventional CT

Clinical evaluation of an image-guided cochlear implant programming strategy

The cochlear implant (CI) has been labeled the most successful neural prosthesis. Despite this success, a significant number of CI recipients experience poor speech understanding, and, even among the best performers, restoration to normal auditory fidelity is rare. While significant research efforts have been devoted to improving stimulation strategies, few developments have led to significant hearing improvement over the past two decades. We have recently introduced image processing techniques that open a new direction for advancement in this field by making it possible, for the first time, to determine the position of implanted CI electrodes relative to the nerves they stimulate using computed tomography images. In this article, we present results of an image-guided, patient-customized approach to stimulation that utilizes the electrode position information our image processing techniques provide. This approach allows us to identify electrodes that cause overlapping stimulation patterns and to deactivate them from a patient's map. This individualized mapping strategy yields significant improvement in speech understanding in both quiet and noise as well as improved spectral resolution in the 68 adult CI recipients studied to date. Our results indicate that image guidance can improve hearing outcomes for many existing CI recipients without requiring additional surgery or the use of 'experimental' stimulation strategies, hardware or software.

Automatic localization of the fetal cerebellum on 3D ultrasound volumes

PURPOSE

Assessment of the fetal cerebellar volume on 3D ultrasound data sets is very important in the clinical evaluation of the fetal growth and health. However, the irregular shape of the cerebellum and the strong artifacts of ultrasound images complicate the segmentation without manual intervention. In this paper, the authors propose an approach to locate the cerebellum automatically, which is considered as a prework of the segmentation.

METHODS

The authors present a weighted Hough transform and a constrained randomized Hough transform to detect the fetal brain midline and the skull, respectively. By combining the location information of these two structures with local image features, a constrained probabilistic boosting tree is then proposed to search the cerebellum.

RESULTS

This algorithm was tested on ultrasound volumes of the fetal head with the gestational age ranging from 20 to 33 weeks. Compared with manual measurements, this algorithm obtained a satisfactory performance with the mean Dice similarity coefficient of 0.92 and the average processing time of 0.75 s per case.

CONCLUSIONS

The results demonstrate that the proposed method is an automatic, fast, and accurate tool for searching the fetal cerebellum on ultrasound volumes.

Automated segmentation of the incus and malleus ossicles in conventional tri-dimensional computed tomography images

This article proposes a fully automated computational solution to segment the incus and malleus ear ossicles in conventional tri-dimensional X-ray computed tomography images. The solution uses a registration-based segmentation paradigm, followed by image segmentation refinement. It was tested against a dataset comprising 21 computed tomography volumetric images of the ear acquired using standard protocols and with resolutions varying from 0.162 × 0.162 × 0.6 to 0.166 × 0.166 × 1.0 mm(3). The images used were randomly selected from subjects who had had a computed tomography examination of the ear due to ear-related pathologies. Dice's coefficient and the Hausdorff distance were used to compare the results of the automated segmentation against those of a manual segmentation performed by two experts. The mean agreement between automated and manual segmentations was equal to 0.956 (Dice's coefficient), and the mean Hausdorff distance among the shapes obtained was 1.14 mm, which is approximately equal to the maximum distance between the neighbouring voxels in the dataset tested. The results confirm that the automated segmentation of the incus and malleus ossicles in tri-dimensional images acquired from patients with ear-related pathologies, using conventional computed tomography scanners and standard protocols, is feasible, robust and accurate. Thus, the solution developed can be employed efficiently in computed tomography ear examinations to help radiologists and otolaryngologists in the evaluation of bi-dimensional slices by providing the related tri-dimensional model.

Analysis of intersubject variations in intracochlear and middle ear surface anatomy for cochlear implantation

HYPOTHESIS

We hypothesize that surface landmarks surrounding the round window typically used to guide electrode placement during cochlear implantation (CI) exhibit substantial variability with respect to intracochlear anatomy.

BACKGROUND

Recent publications suggest that both atraumatic electrode insertion and electrode location within the scala tympani can affect auditory performance after CI. However, current techniques for electrode insertion rely on surface landmarks alone for navigation, without actual visualization of intracochlear structures other than what can be seen through a surgically created cochleostomy. In this study, we quantify how well the position of intracochlear anatomy is predicted by surface landmarks surrounding the round window.

METHODS

Structures representing middle ear surface and intracochlear anatomy were reconstructed in μCT scans of 10 temporal bone specimens. These structures were then reoriented into a normalized coordinate system to facilitate measurement of inter-subject anatomical shape variations.

RESULTS

Only minor intersubject variations were detected for intracochlear anatomy (maximum deviation, 0.71 mm; standard deviation, 0.21 mm), with greatest differences existing near the hook and apex. Larger intersubject variations in intracochlear structures were detected when considered relative to surface landmarks surrounding the round window (maximum deviation, 0.83 mm; standard deviation, 0.54 mm).

CONCLUSION

The cochlea and its scala exhibit considerable variability in relation to middle ear surface landmarks. While support for more precise, atraumatic CI electrode insertion techniques is growing in the otologic community, landmark guided insertion techniques have limited precision. Refining the CI insertion process may require the development of image-guidance systems for use in otologic surgery.

Impact of electrode design and surgical approach on scalar location and cochlear implant outcomes

OBJECTIVES/HYPOTHESIS

Three surgical approaches: cochleostomy (C), round window (RW), and extended round window (ERW); and two electrodes types: lateral wall (LW) and perimodiolar (PM), account for the vast majority of cochlear implantations. The goal of this study was to analyze the relationship between surgical approach and electrode type with final intracochlear position of the electrode array and subsequent hearing outcomes.

STUDY DESIGN

Comparative longitudinal study.

METHODS

One hundred postlingually implanted adult patients were enrolled in the study. From the postoperative scan, intracochlear electrode location was determined and using rigid registration, transformed back to the preoperative computed tomography which had intracochlear anatomy (scala tympani and scala vestibuli) specified using a statistical shape model based on 10 microCT scans of human cadaveric cochleae. Likelihood ratio chi-square statistics were used to evaluate for differences in electrode placement with respect to surgical approach (C, RW, ERW) and type of electrode (LW, PM).

RESULTS

Electrode placement completely within the scala tympani (ST) was more common for LW than were PM designs (89% vs. 58%; P < 0.001). RW and ERW approaches were associated with lower rates of electrode placement outside the ST than was the cochleostomy approach (9%, 16%, and 63%, respectively; P < 0.001). This pattern held true regardless of whether the implant was LW or PM. When examining electrode placement and hearing outcome, those with electrode residing completely within the ST had better consonant-nucleus-consonant word scores than did patients with any number of electrodes located outside the ST (P = 0.045).

CONCLUSION

These data suggest that RW and ERW approaches and LW electrodes are associated with an increased likelihood of successful ST placement. Furthermore, electrode position entirely within the ST confers superior audiological outcomes.

LEVEL OF EVIDENCE

2b.

An artifact-robust, shape library-based algorithm for automatic segmentation of inner ear anatomy in post-cochlear-implantation CT

A cochlear implant (CI) is a device that restores hearing using an electrode array that is surgically placed in the cochlea. After implantation, the CI is programmed to attempt to optimize hearing outcome. Currently, we are testing an image-guided CI programming (IGCIP) technique we recently developed that relies on knowledge of relative position of intracochlear anatomy to implanted electrodes. IGCIP is enabled by a number of algorithms we developed that permit determining the positions of electrodes relative to intra-cochlear anatomy using a pre- and a post-implantation CT. One issue with this technique is that it cannot be used for many subjects for whom a pre-implantation CT was not acquired. Pre-implantation CT has been necessary because it is difficult to localize the intra-cochlear structures in post-implantation CTs alone due to the image artifacts that obscure the cochlea. In this work, we present an algorithm for automatically segmenting intra-cochlear anatomy in post-implantation CTs. Our approach is to first identify the labyrinth and then use its position as a landmark to localize the intra-cochlea anatomy. Specifically, we identify the labyrinth by first approximately estimating its position by mapping a labyrinth surface of another subject that is selected from a library of such surfaces and then refining this estimate by a standard shape model-based segmentation method. We tested our approach on 10 ears and achieved overall mean and maximum errors of 0.209 and 0.98 mm, respectively. This result suggests that our approach is accurate enough for developing IGCIP strategies based solely on post-implantation CTs.

Automatic segmentation of intra-cochlear anatomy in post-implantation CT of unilateral cochlear implant recipients

A cochlear implant (CI) is a neural prosthetic device that restores hearing by directly stimulating the auditory nerve using an electrode array that is implanted in the cochlea. In CI surgery, the surgeon accesses the cochlea and makes an opening where he/she inserts the electrode array blind to internal structures of the cochlea. Because of this, the final position of the electrode array relative to intra-cochlear anatomy is generally unknown. We have recently developed an approach for determining electrode array position relative to intra-cochlear anatomy using a pre- and a post-implantation CT. The approach is to segment the intra-cochlear anatomy in the pre-implantation CT, localize the electrodes in the post-implantation CT, and register the two CTs to determine relative electrode array position information. Currently, we are using this approach to develop a CI programming technique that uses patient-specific spatial information to create patient-customized sound processing strategies. However, this technique cannot be used for many CI users because it requires a pre-implantation CT that is not always acquired prior to implantation. In this study, we propose a method for automatic segmentation of intra-cochlear anatomy in post-implantation CT of unilateral recipients, thus eliminating the need for pre-implantation CTs in this population. The method is to segment the intra-cochlear anatomy in the implanted ear using information extracted from the normal contralateral ear and to exploit the intra-subject symmetry in cochlear anatomy across ears. To validate our method, we performed experiments on 30 ears for which both a pre- and a post-implantation CT are available. The mean and the maximum segmentation errors are 0.224 and 0.734mm, respectively. These results indicate that our automatic segmentation method is accurate enough for developing patient-customized CI sound processing strategies for unilateral CI recipients using a post-implantation CT alone.

Minimally invasive image-guided cochlear implantation surgery: first report of clinical implementation

OBJECTIVES/HYPOTHESIS

Minimally invasive image-guided approach to cochlear implantation (CI) involves drilling a narrow, linear tunnel to the cochlea. Reported herein is the first clinical implementation of this approach.

STUDY DESIGN

Prospective cohort study.

METHODS

On preoperative computed tomography (CT), a safe linear trajectory through the facial recess targeting the scala tympani was planned. Intraoperatively, fiducial markers were bone-implanted, a second CT was acquired, and the trajectory was transferred from preoperative to intraoperative CT. A customized microstereotactic frame was rapidly designed and constructed to constrain a surgical drill along the desired trajectory. Following sterilization, the frame was employed to drill the tunnel to the middle ear. After lifting a tympanomeatal flap and performing a cochleostomy, the electrode array was threaded through the drilled tunnel and into the cochlea.

RESULTS

Eight of nine patients were successfully implanted using the proposed approach with six insertions completely within the scala tympani. Traditional mastoidectomy was performed on one patient following difficulty threading the electrode array via the narrow tunnel. Other difficulties encountered included use of the backup implant when an electrode was dislodged during threading via the tunnel, tip fold-over, and facial nerve paresis (House-Brackmann II/VI at 12 months) secondary to heat during drilling. The average time of intervention was 182 ± 36 minutes.

CONCLUSIONS

Minimally invasive image-guided CI is clinically achievable. Further clinical study is necessary to address technological difficulties during drilling and insertion, and to assess potential benefits including decreased time of intervention, standardization of surgical intervention, and decreased tissue dissection potentially leading to shorter recovery and earlier implant activation.

Image-guidance enables new methods for customizing cochlear implant stimulation strategies

Over the last 20 years, cochlear implants (CIs) have become what is arguably the most successful neural prosthesis to date. Despite this success, a significant number of CI recipients experience marginal hearing restoration, and, even among the best performers, restoration to normal fidelity is rare. In this paper, we present image processing techniques that can be used to detect, for the first time, the positions of implanted CI electrodes and the nerves they stimulate for individual CI users. These techniques permit development of new, customized CI stimulation strategies. We present one such strategy and show that it leads to significant hearing improvement in an experiment conducted with 11 CI recipients. These results indicate that image-guidance can be used to improve hearing outcomes for many existing CI recipients without requiring additional surgical procedures.

Automatic pre- to intra-operative CT registration for image-guided cochlear implant surgery

Percutaneous cochlear implantation (PCI) is a minimally-invasive image-guided cochlear implant approach, where access to the cochlea is achieved by drilling a linear channel from the skull surface to the cochlea. The PCI approach requires pre- and intra-operative planning. Computation of a safe linear drilling trajectory is performed in a preoperative CT. This trajectory is mapped to intraoperative space using the transformation matrix that registers the pre- and intra-operative CTs. However, the difference in orientation between the pre- and intra-operative CTs is too extreme to be recovered by standard, gradient descent-based registration methods. Thus far, the registration has been initialized manually by an expert. In this paper, we present a method that aligns the scans completely automatically. We compared the performance of the automatic approach to the registration approach when an expert does the manual initialization on 11 pairs of scans. There is a maximum difference of 0.18 mm between the entry and target points of the trajectory mapped with expert initialization and the automatic registration method. This suggests that the automatic registration method is accurate enough to be used in a PCI surgery.

A manually operated, advance off-stylet insertion tool for minimally invasive cochlear implantation surgery

The current technique for cochlear implantation (CI) surgery requires a mastoidectomy to gain access to the cochlea for electrode array insertion. It has been shown that microstereotactic frames can enable an image-guided, minimally invasive approach to CI surgery called percutaneous cochlear implantation (PCI) that uses a single drill hole for electrode array insertion, avoiding a more invasive mastoidectomy. Current clinical methods for electrode array insertion are not compatible with PCI surgery because they require a mastoidectomy to access the cochlea; thus, we have developed a manually operated electrode array insertion tool that can be deployed through a PCI drill hole. The tool can be adjusted using a preoperative CT scan for accurate execution of the advance off-stylet (AOS) insertion technique and requires less skill to operate than is currently required to implant electrode arrays. We performed three cadaver insertion experiments using the AOS technique and determined that all insertions were successful using CT and microdissection.

Comparison of cochlear implant relevant anatomy in children versus adults

HYPOTHESIS

To test whether there are significant differences in pediatric and adult temporal bone anatomy as related to cochlear implant (CI) surgery.

BACKGROUND

Surgeons rely upon anatomic landmarks including the round window (RW) and facial recess (FR) to place CI electrodes within the scala tympani. Anecdotally, clinicians report differences in orientation of such structures in children versus adults.

METHODS

Institutional review board approval was obtained. High-resolution computed tomographic scans of 24 pediatric patients (46 ears) and 20 adult patients (40 ears) were evaluated using software consisting of a model-based segmentation algorithm that automatically localizes and segments temporal bone anatomy (e.g., facial nerve, chorda tympani, external auditory canal [EAC], and cochlea). On these scans, angles pertinent anatomy were manually delineated and measured blinded as to the age of the patient.

RESULTS

The EAC and FR were more parallel to the basal turn (BT) of the cochlea in children versus adults ([symbol in text] EAC:BT 20.55 degrees versus 24.28 degrees, p = 0.003; [symbol in text] FR:BT 5.15 degrees versus 6.88 degrees, p = 0.009). The RW was more closely aligned with the FR in children versus adults ([symbol in text] FR:RW 30.43 degrees versus 36.67 degrees, p = 0.009). Comparing the lateral portion of the EAC (using LatEAC as a marker) to the most medial portion (using [symbol in text] TM as a marker), the measured angle was 136.57 degrees in children and 172.20 degrees in adults (p < 0.001).

CONCLUSION

There are significant differences in the temporal bone anatomy of children versus adults pertinent to CI electrode insertion.

Variability of the temporal bone surface's topography: implications for otologic surgery

Otologic surgery is performed for a variety of reasons including treatment of recurrent ear infections, alleviation of dizziness, and restoration of hearing loss. A typical ear surgery consists of a tympanomastoidectomy in which both the middle ear is explored via a tympanic membrane flap and the bone behind the ear is removed via mastoidectomy to treat disease and/or provide additional access. The mastoid dissection is performed using a high-speed drill to excavate bone based on a pre-operative CT scan. Intraoperatively, the surface of the mastoid component of the temporal bone provides visual feedback allowing the surgeon to guide their dissection. Dissection begins in "safe areas" which, based on surface topography, are believed to be correlated with greatest distance from surface to vital anatomy thus decreasing the chance of injury to the brain, large blood vessels (e.g. the internal jugular vein and internal carotid artery), the inner ear, and the facial nerve. "Safe areas" have been identified based on surgical experience with no identifiable studies showing correlation of the surface with subsurface anatomy. The purpose of our study was to investigate whether such a correlation exists. Through a three-step registration process, we defined a correspondence between each of twenty five clinically-applicable temporal bone CT scans of patients and an atlas and explored displacement and angular differences of surface topography and depth of critical structures from the surface of the skull. The results of this study reflect current knowledge of osteogenesis and anatomy. Based on two features (distance and angular difference), two regions (suprahelical and posterior) of the temporal bone show the least variability between surface and subsurface anatomy.

Statistical shape model segmentation and frequency mapping of cochlear implant stimulation targets in CT

Cochlear implant (CI) surgery is considered standard of care treatment for severe hearing loss. CIs are currently programmed using a one-size-fits-all type approach. Individualized, position-based CI programming schemes have the potential to significantly improve hearing outcomes. This has not been possible because the position of stimulation targets is unknown due to their small size and lack of contrast in CT. In this work, we present a segmentation approach that relies on a weighted active shape model created using microCT scans of the cochlea acquired ex-vivo in which stimulation targets are visible. The model is fitted to the partial information available in the conventional CTs and used to estimate the position of structures not visible in these images. Quantitative evaluation of our method results in Dice scores averaging 0.77 and average surface errors of 0.15 mm. These results suggest that our approach can be used for position-dependent image-guided CI programming methods.