Deep reinforcement learning and convolutional autoencoders for anomaly detection of congenital inner ear malformations in clinical CT images.

Detection of abnormalities within the inner ear is a challenging task even for experienced clinicians. In this study, we propose an automated method for automatic abnormality detection to provide support for the diagnosis and clinical management of various otological disorders. We propose a framework for inner ear abnormality detection based on deep reinforcement learning for landmark detection which is trained uniquely in normative data. In our approach, we derive two abnormality measurements: Dimage and Uimage. The first measurement, Dimage, is based on the variability of the predicted configuration of a well-defined set of landmarks in a subspace formed by the point distribution model of the location of those landmarks in normative data. We create this subspace using Procrustes shape alignment and Principal Component Analysis projection. The second measurement, Uimage, represents the degree of hesitation of the agents when approaching the final location of the landmarks and is based on the distribution of the predicted Q-values of the model for the last ten states. Finally, we unify these measurements in a combined anomaly measurement called Cimage. We compare our method's performance with a 3D convolutional autoencoder technique for abnormality detection using the patch-based mean squared error between the original and the generated image as a basis for classifying abnormal versus normal anatomies. We compare both approaches and show that our method, based on deep reinforcement learning, shows better detection performance for abnormal anatomies on both an artificial and a real clinical CT dataset of various inner ear malformations with an increase of 11.2% of the area under the ROC curve. Our method also shows more robustness against the heterogeneous quality of the images in our dataset.

Adult Cochlear Implantation Under Local Anesthesia and Conscious Sedation with Dexmedetomidine: Efficacy and a Method to Interact with the Conscious and Cooperative Patient.

BACKGROUND: This study describes the efficacy of cochlear implantation under local anesthesia with conscious sedation with dexmedetomi- dine in adult patients and proposes a method to communicate with the conscious and cooperative patient intraoperatively. This less invasive anesthetic procedure is suitable for patients with comorbidities preventing general anesthesia. METHODS: Unilateral cochlear implantation with Oticon Medical systems was performed in 10 adult patients with comorbidities preventing general anesthesia. Classical cochlear implantation was performed under local anesthesia and conscious sedation with dexmedetomidine. Cue cards were used to support intraoperative dialogue. Outcome measures were intraoperative adverse events, patient perceptions, as well as post- operative completions measured with a questionnaire. RESULTS: The procedure was successful for all 10 patients. Dexmedetomidine lead to rapid and successful conscious sedation and no case of high blood pressure or aggravation of comorbidities was noted. Stapedial reflex measurements led to reliable thresholds. The usage of the cue cards was successful: patients were able to read the cue cards and thereby the medical team could inform the patients of surgical progress and ask the patients questions. CONCLUSION: Cochlear implantation and intraoperative dialogue with the conscious and cooperative patient is possible. The main advantage of the anesthetic procedure is the reduction in intra- and postoperative complications. Further, expected benefits include a less invasive procedure, the conscious state of the patient which enables the recording of auditory perception, and the absence of nonauditory percepts such as facial nerve stimulation during implant stimulation, a shorter surgical duration, and lower-associated costs.

Spontaneous bone bed formation in pediatric cochlear implantation is associated with duration of implantation.

OBJECTIVES: This study investigated the long-term postoperative spontaneous formation of a bone bed in pediatric cochlear implant patients for whom no bone bed was drilled during the surgery. METHODS: A cross-sectional observational study of skull thickness under and on the edges of the cochlear implant receiver/stimulator in children with computed tomography (CT scan) ≥6 months after implantation was performed. In total, 37 pediatric patients from a single tertiary center underwent cochlear implantation without bone bed drilling and with screw fixation of the receiver/stimulator. RESULTS: The patients were on average 36.2 ± 20.5 months at implantation (range 8-96 months). At the time of the CT scan, the average duration of implantation was 25.3 ± 17.9 months (range 6-91 months). The average depth of the bone bed that formed spontaneously since implantation was 1.83 ± 0.39 mm (range 0.39-3.04 mm). Linear regression identified that the depth of the bone bed increased significantly with duration of implantation (ß = 0.389, p = 0.009), but age at implantation was not associated with bone bed depth. CONCLUSIONS: A spontaneously formed temporal bone bed was observed in pediatric CI patients already six months after implantation. A deeper bone bed was measured in children who have had their CI for a longer period. A spontaneously formed bone bed is likely to combine the benefits of a surgically drilled bone bed, whilst limiting the duration of the surgery and thereby associated costs.