Nationwide Prevalence of Video Head Impulse Test Abnormality and its Risk Factors in South Korea.

OBJECTIVE: To identify the prevalence of and relevant information for video head impulse test (vHIT) abnormality in a large population. STUDY DESIGN: A cross-sectional design. SETTING: Korean National Health and Nutrition Examination Survey, 2021. METHODS: The sample was representative of the Korean population, with 2237 participants aged ≥40 years. A vHIT was performed to evaluate vestibular function. The vestibulo-ocular reflex (VOR) gain and the presence of reproducible catch-up saccades was assessed in a vHIT. Participants also completed questionnaires for demographics, socioeconomic status, and basic information regarding systemic diseases and dizziness and underwent hearing tests with automated pure-tone audiometry. RESULTS: The prevalence of vHIT abnormality was 22.5%, with unilateral (14.3%) being more common than bilateral (8.2%). The prevalence of vHIT abnormality increased significantly with age, with the highest rate observed in individuals aged >70 years (42.5%). Both hearing and VOR gain deteriorated with age, but the patterns of age-related progression were different. While hearing loss (HL) deteriorated gradually and progressively throughout adulthood, VOR gain deterioration was markedly evident after 70 years of age. CONCLUSION: Considering the high prevalence of vHIT abnormality, appropriate social and medical policies are needed to prevent associated injuries and improve patients' quality of life. The distinct age-related changes in HL and objective findings of vestibular dysfunction indicate the need for different approaches to address these social problems in aging countries.

Characterization of Vestibular Phenotypes in Patients with Genetic Hearing Loss.

Background: The vestibular phenotypes of patients with genetic hearing loss are poorly understood. Methods: we performed genetic testing including exome sequencing and vestibular function tests to investigate vestibular phenotypes and functions in patients with genetic hearing loss. Results: Among 627 patients, 143 (22.8%) had vestibular symptoms. Genetic variations were confirmed in 45 (31.5%) of the 143 patients. Nineteen deafness genes were linked with vestibular symptoms; the most frequent genes in autosomal dominant and recessive individuals were COCH and SLC26A4, respectively. Vestibular symptoms were mostly of the vertigo type, recurrent, and persisted for hours in the genetically confirmed and unconfirmed groups. Decreased vestibular function in the caloric test, video head impulse test, cervical vestibular-evoked myogenic potential, and ocular vestibular-evoked myogenic potential was observed in 42.0%, 16.3%, 57.8%, and 85.0% of the patients, respectively. The caloric test revealed a significantly higher incidence of abnormal results in autosomal recessive individuals than in autosomal dominant individuals (p = 0.011). The genes, including SLC26A4, COCH, KCNQ4, MYH9, NLRP3, EYA4, MYO7A, MYO15A, and MYH9, were heterogeneously associated with abnormalities in the vestibular function test. Conclusions: In conclusion, diverse vestibular symptoms are commonly concomitant with genetic hearing loss and are easily overlooked.

Vestibular dysfunction in lateral semicircular canal dysplasia.

Introduction: Lateral semicircular canal (LSCC) dysplasia is the most common inner ear malformation. The severity of dysplasia can appear in various spectrums, from a short and broad LSCC with normal or small-sized central bony island (CBI) to a single fluid-filled cavity confluent with the vestibule without CBI. However, reports on the association between LSCC dysplasia and the loss of vestibular function are still lacking. In this study, the results of vestibular function tests [caloric test and video-head impulse test (vHIT)] in patients with LSCC dysplasia were analyzed and compared between groups with and without CBI. Methods: This study retrospectively enrolled 17 patients (23 ears) who had LSCC dysplasia following computed tomography or magnetic resonance imaging and underwent vestibular function tests. Results: LSCC dysplasia was observed unilaterally in 11 patients and bilaterally in six patients. Nine of 23 ears had CBIs, and 14 ears had no CBI. Three of 17 patients experienced dizziness. Abnormal caloric tests were detected in 11 of the 16 patients who underwent the caloric tests (69%); in contrast, 11 of 12 patients who underwent the vHIT (92%) had normal LSCC vestibulo-ocular reflex (VOR) gain on vHIT. A significant correlation was found between the maximum slow-phase velocity of the caloric test and LSCC VOR gain of the vHIT (correlation coefficient 0.792, p = 0.004). The CBI-absent group showed significantly lower SPV and LSCC VOR gains than the CBI-present group (p = 0.001 and 0.004, respectively). Discussion: LSCC dysplasia impairs VOR function, especially in the absence of CBI.

Results of Eustachian tube balloon dilation measured using the nine-step test.

Suggested several decades ago, the nine-step test is an intuitive test of Eustachian tube function. However, studies employing the nine-step test to assess the results of Eustachian tube balloon dilation (EBD) are limited. We aimed to objectively evaluate the efficacy of EBD in opening failure patients with decreased maximal peak pressure difference (MPD) using the nine-step test. Patients who had MPD values ≤ 13 daPa in the nine-step test were enrolled. The patients were categorized into two groups according to treatment decisions after discussion with a clinician: an EBD group (N = 26) and a medication group (N = 30). One month after treatment, the seven-item Eustachian Tube Dysfunction Questionnaire (ETDQ7) and the nine-step test were administered to all participants and subgroups of symptomatic participants (ETDQ7 > 15). MPD improved (increased) in both the EBD group and the medication group. ETDQ7 values improved (decreased) in the EBD group, but not in the medication group. In subgroup analysis, MPD and ETDQ7 values improved only in the symptomatic EBD group. According to the nine-step test, EBD can normalize 53.8% of decreased MPD. Posttreatment MPD and ETDQ7 scores were significantly better in the EBD group than in the medication group. However, EBD in patients with abnormal nine-step test results seemed less efficacious when the treatment results of the medication group were considered.

Human-machine cooperation meta-model for clinical diagnosis by adaptation to human expert's diagnostic characteristics.

Artificial intelligence (AI) using deep learning approaches the capabilities of human experts in medical image diagnosis. However, due to liability issues in medical decisions, AI is often relegated to an assistant role. Based on this responsibility constraint, the effective use of AI to assist human intelligence in real-world clinics remains a challenge. Given the significant inter-individual variations in clinical decisions among physicians based on their expertise, AI needs to adapt to individual experts, complementing weaknesses and enhancing strengths. For this adaptation, AI should not only acquire domain knowledge but also understand the specific human experts it assists. This study introduces a meta-model for human-machine cooperation that first evaluates each expert's class-specific diagnostic tendencies using conditional probability, based on which the meta-model adjusts the AI's predictions. This meta-model was applied to ear disease diagnosis using otoendoscopy, highlighting improved performance when incorporating individual diagnostic characteristics, even with limited evaluation data. The highest accuracy was achieved by combining each expert's conditional probabilities with machine classification probability, using optimal weights specific to each individual's overall classification accuracy. This tailored model aims to mitigate potential misjudgments due to psychological effects caused by machine suggestions and to capitalize on the unique expertise of individual clinicians.

Diagnostic Value of the Air-Bone Threshold Gap in Stapes Fixation.

INTRODUCTION: The Carhart notch is a well-known sign of stapes fixation. However, previous studies have reported that the Carhart notch is not specific to stapes fixation and is also present in other middle ear diseases. Therefore, this study investigated the diagnostic value of threshold gap between air conduction and bone conduction (ABG) for stapes fixation, instead of the bone conduction dip representing the Carhart notch. METHODS: A total of 199 ears that underwent exploratory tympanotomy were enrolled in this retrospective study. They were categorized into three groups according to surgical findings: stapes fixation (SF), other ossicle fixation (OF), and chain disconnection (CD). Preoperative pure-tone audiograms and impedance audiograms were compared between the groups. RESULTS: The incidence of the Carhart notch did not differ between the groups. The ABG at 2,000 Hz showed a good diagnostic performance for distinguishing between the SF and CD groups (area under the curve, AUC = 0.816, p < 0.001), but poor performance for distinguishing between the SF and OF groups (AUC = 0.662, p = 0.003). Bone conduction at 2,000 Hz showed a moderate performance for distinguishing between the SF and CD groups (AUC = 0.707, p < 0.001) and did not show statistically significant results for distinguishing between the SF and OF groups (AUC = 0.594, p = 0.080). The tympanic membrane compliance was significantly higher in the CD group than in the SF group (p = 0.001). CONCLUSIONS: The Carhart notch was not a specific finding of SF. The sensitivity and specificity of ABG ≤15 dB at 2,000 Hz for distinguishing between SF and CD were 60.4% and 89.2%, respectively. To prepare for surgical interventions in patients with conductive hearing loss but a normal tympanic membrane, clinicians should comprehensively consider these results.

Comprehensive Prediction Model, Including Genetic Testing, for the Outcomes of Cochlear Implantation.

OBJECTIVES: Despite growing interest in the genetic contribution to cochlear implant (CI) outcomes, only a few studies with limited samples have examined the association of CI outcomes with genetic etiologies. We analyzed CI outcomes using known predictors and genetic testing results to obtain a comprehensive understanding of the impact of genetic etiologies. DESIGN: We retrospectively reviewed the medical records and images of patients who underwent cochlear implantation and genetic testing at a single tertiary medical institution, between May 2008 and December 2020. After excluding those whose speech test results were unavailable, and those in whom the implant was removed due to complications, such as infection or device failure, 203 patients were included in this study. The participants were categorized into adult (≥19 years), child (2-18 years), and infant (<24 months) groups. Outcomes were measured based on categories of auditory perception, monosyllable, disyllable, and sentence scores. For the infant group, the Infant-Toddler Meaningful Auditory Integration Scale score was used. RESULTS: Among the 203 participants, a causative genetic variant was identified in 117 (57.6%) individuals. The presence of a causative variant was significantly associated with better CI outcomes in the infant group (ß = 0.23; 95% confidence interval, 0.01 to 0.47; p = 0.044), but not in the child and adult groups. In the genetically confirmed patients without cochlear malformation, genetic variants involving the spiral ganglion was a poor prognostic factor in the child group (ß = -57.24; 95% confidence interval, -90.63 to -23.75; p = 0.004). CONCLUSIONS: The presence of known genetic etiology of hearing loss was associated with better CI outcomes in the infant group, but not in the child and adult groups. A neural-type genetic variant was a poor prognostic factor in the genetically diagnosed child subgroup without cochlear malformation. Careful genetic counseling should be performed before cochlear implantation.

High incidence of cleft palate and vomer deformities in patients with Eustachian tube dysfunction.

Although the cleft palate is regarded as a contraindication for Eustachian tube ballooning, the presence of submucosal cleft palate may be overlooked while diagnosing Eustachian tube dysfunction. Therefore, we aimed to determine the incidence of the presence of a hard palate bony notch and vomer defect, which indicate the presence of submucosal cleft palate in patients with Eustachian tube dysfunction. In the Eustachian tube dysfunction group (n = 28), 4 patients (14.3%) exhibited a hard palate bony notch and a concurrent vomer defect. Three of them exhibited the presence of occult submucosal cleft palate, which had not been diagnosed previously. None of the control group (n = 39) showed any of these findings. The hard palate length of patients in the Eustachian tube dysfunction group was significantly lesser than that of those in the control group (34.2 ± 5.6 mm vs. 37.2 ± 2.1 mm, P = 0.016). Patients with Eustachian tube dysfunction have a high incidence of submucosal cleft palate and its occult variant, which are challenging to diagnose without any preexisting suspicion. Clinicians should evaluate the hard palate and vomer to exclude the presence of occult submucosal cleft palate while diagnosing Eustachian tube dysfunction.

Saccular Pathology Is Most Commonly Found in Patients With General Vestibular Disorders.

OBJECTIVES: The first purpose of this study was to investigate the difference in the frequency of involvement of the superior vestibular nerve (SVN) and inferior vestibular nerve (IVN) territories in general vestibular disorders, and to identify which IVN territory was more commonly involved in patients with IVN lesions. The second purpose was to investigate the correlation of the degree of each saccular and posterior semicircular canal (PSCC) dysfunction, as represented by the parameters of cervical vestibular evoked myogenic potential (cVEMP) and video head impulse test (vHIT), in patients with pathology of the IVN territory. METHODS: In total, 346 patients with dizziness who underwent the caloric test, cVEMP, and vHIT were enrolled. Canal weakness in the caloric test, interaural amplitude difference (IAD) of cVEMP, and vestibulo-ocular reflex gain of the vestibulo-ocular reflex gain of the posterior semicircular canal (p-VOR) in vHIT were analyzed. RESULTS: Among the enrolled patients, 15.6% had total vestibular nerve dysfunction, 14.5% had solely SVN dysfunction, and 29.5% had solely IVN dysfunction. Isolated saccular pathology was most common in patients with IVN pathology, followed by those with total IVN dysfunction and PSCC dysfunction. IAD and p-VOR were statistically well correlated, and the correlation was strongest in patients with both pathologic IAD and pathologic p-VOR (n=23, r=0.944), followed by patients with normal IAD and pathologic p-VOR (n=27, r=0.762) and patients with pathologic IAD and normal p-VOR (n=106, r=0.339). CONCLUSION: Abnormal results were more common in vestibular tests investigating the IVN than in vestibular tests investigating the SVN in patients with general vestibular disorders. Isolated saccular pathology was more frequent than PSCC or combined pathology in patients with IVN dysfunction. Patients with abnormal p-VOR in vHIT had a higher probability of having both saccular and PSCC pathologies than patients with an abnormal IAD. This study describes the characteristics of vestibular-system subregions and provides guidance for clinically interpreting the combination of cVEMP and vHIT results.

Purinergic signaling in the peripheral vestibular system.

The inner ear comprises the cochlea and vestibular system, which detect sound and acceleration stimulation, respectively. The function of the inner ear is regulated by ion transport activity among sensory epithelial cells, neuronal cells, non-sensory epithelial cells, and luminal fluid with a unique ionic composition of high [K+] and low [Na+], which enables normal hearing and balance maintenance. One of the important mechanisms regulating ion transport in the inner ear is purinergic signaling. Various purinergic receptors are distributed throughout inner ear epithelial cells and neuronal cells. To date, most studies have focused on the role of purinergic receptors in the cochlea, and few studies have examined these receptors in the vestibular system. As purinergic receptors play an important role in the cochlea, they would likely do the same in the vestibular system, which is fairly similar to the cochlea in cellular structure and function. Based on available studies performed to date, purinergic signaling is postulated to be involved in the regulation of ion homeostasis, protection of hair cells, otoconia formation, and regulation of electrical signaling from the sensory epithelium to vestibular neurons. In this review, the distribution and roles of purinergic receptors in the peripheral vestibular system are summarized and discussed.

A Case of Isolated Congenital Stapedial Suprastructure Fixation.

We report a case of conductive hearing loss caused by isolated congenital stapedial suprastructure fixation with normal footplate mobility. A 60-year-old woman visited the clinic for right-sided mixed hearing loss. Exploratory tympanotomy revealed a bony synostosis between the stapedial suprastructure and promontory, while all the ossicles were present and normally shaped. As the bony synostosis was separated, the stapes became mobile. This is the first report in the medical literature of this congenital ear anomaly. This case also illustrates that stapedial fixation can occur in the suprastructure as well as in the footplate; thus, one must be mindful of this when performing exploratory tympanotomy for stapes fixation.

Effect of Dexamethasone Combination with Gentamicin in Chemical Labyrinthectomy on Hearing Preservation and Vertigo Control in Patients with Unilateral Meniere's Disease: A Randomized Controlled Clinical Trial.

Chemical labyrinthectomy using gentamicin is a popular method for treating intractable vertigo attacks in Meniere's disease. However, the risk of hearing loss remains a major concern for clinicians. We investigated the effect of simultaneous dexamethasone and gentamicin application on hearing preservation and vertigo control in patients with intractable unilateral Meniere's disease. A single-institutional, prospective, single-blinded, randomized clinical trial was conducted. Gentamicin-soaked Gelfoam® was directly applied on the oval window following middle ear exploration. On the round window, dexamethasone-soaked Gelfoam® was applied in the gentamicin with dexamethasone group (GD group, n = 18), and saline-soaked Gelfoam® was applied in the gentamicin with sham reagent group (GO group, n = 19). The hearing change 8 weeks after the procedure and vertigo control 2-12 months after the procedure were investigated. The high-frequency hearing threshold was significantly increased in the GO group (p = 0.005 and 0.012 for 4 and 8 kHz, respectively), but not in the GD group. The short-term (2-6 months) vertigo control was more successful in the GD group (57.89% vs. 94.44%, p = 0.019), but long-term control (6-12 months) was insignificant. In conclusion, the combined application of gentamicin and dexamethasone in chemical labyrinthectomy is an effective method for protecting high-frequency hearing and vertigo control.

Differential Biases and Variabilities of Deep Learning-Based Artificial Intelligence and Human Experts in Clinical Diagnosis: Retrospective Cohort and Survey Study.

BACKGROUND: Deep learning (DL)-based artificial intelligence may have different diagnostic characteristics than human experts in medical diagnosis. As a data-driven knowledge system, heterogeneous population incidence in the clinical world is considered to cause more bias to DL than clinicians. Conversely, by experiencing limited numbers of cases, human experts may exhibit large interindividual variability. Thus, understanding how the 2 groups classify given data differently is an essential step for the cooperative usage of DL in clinical application. OBJECTIVE: This study aimed to evaluate and compare the differential effects of clinical experience in otoendoscopic image diagnosis in both computers and physicians exemplified by the class imbalance problem and guide clinicians when utilizing decision support systems. METHODS: We used digital otoendoscopic images of patients who visited the outpatient clinic in the Department of Otorhinolaryngology at Severance Hospital, Seoul, South Korea, from January 2013 to June 2019, for a total of 22,707 otoendoscopic images. We excluded similar images, and 7500 otoendoscopic images were selected for labeling. We built a DL-based image classification model to classify the given image into 6 disease categories. Two test sets of 300 images were populated: balanced and imbalanced test sets. We included 14 clinicians (otolaryngologists and nonotolaryngology specialists including general practitioners) and 13 DL-based models. We used accuracy (overall and per-class) and kappa statistics to compare the results of individual physicians and the ML models. RESULTS: Our ML models had consistently high accuracies (balanced test set: mean 77.14%, SD 1.83%; imbalanced test set: mean 82.03%, SD 3.06%), equivalent to those of otolaryngologists (balanced: mean 71.17%, SD 3.37%; imbalanced: mean 72.84%, SD 6.41%) and far better than those of nonotolaryngologists (balanced: mean 45.63%, SD 7.89%; imbalanced: mean 44.08%, SD 15.83%). However, ML models suffered from class imbalance problems (balanced test set: mean 77.14%, SD 1.83%; imbalanced test set: mean 82.03%, SD 3.06%). This was mitigated by data augmentation, particularly for low incidence classes, but rare disease classes still had low per-class accuracies. Human physicians, despite being less affected by prevalence, showed high interphysician variability (ML models: kappa=0.83, SD 0.02; otolaryngologists: kappa=0.60, SD 0.07). CONCLUSIONS: Even though ML models deliver excellent performance in classifying ear disease, physicians and ML models have their own strengths. ML models have consistent and high accuracy while considering only the given image and show bias toward prevalence, whereas human physicians have varying performance but do not show bias toward prevalence and may also consider extra information that is not images. To deliver the best patient care in the shortage of otolaryngologists, our ML model can serve a cooperative role for clinicians with diverse expertise, as long as it is kept in mind that models consider only images and could be biased toward prevalent diseases even after data augmentation.

One-stage cochlear implantation in patients with chronic otitis media using canal wall up mastoidectomy.

BACKGROUND: Cochlear implantation (CI) with subtotal petrosectomy was recommended to avoid the complications for patients with chronic otitis media (COM). OBJECTIVES: To evaluate the surgical outcomes of CI in patients with COM using a one-stage operation with canal wall up mastoidectomy (CWUM). METHODS: Thirty-five patients with COM who underwent CI with CWUM as a one-stage between 2009 and 2017 were participated. They divided into those with inactive COM and active COM. The anatomical success rate, postoperative complication, and hearing outcomes were analyzed. RESULTS: Twenty-four patients had inactive COM and seven with active COM. Three of the 31 patients (9.7%) had otorrhea from the ear undergone surgery. Two of these three patients had myringitis after CI and their symptoms improved after conservative management. Although infection of the tympanic membrane in the third patient was controlled after conservative management, a perforation was left. Postoperative otorrhea occurred in two patients (8.3%) in the inactive COM group and one patient (14.3%) in the active COM group. CONCLUSION: This study indicates that infection control and successful implantation can be achieved through a one-stage CI operation with CWUM in selected patients with COM.HIGHLIGHTSAlthough COM was once considered a contraindication to CI, CI in patients with COM has been made feasible by STP before CI or simultaneously with CI.Simultaneous CI with CWUM was performed for 31 patients with COM.Three patients (9.7%) had minor complications after the surgery and only one patient experienced device explantation which was due to device failure.

Effects of endogenous and exogenous oestrogen exposure on hearing level in postmenopausal women: A cross-sectional study.

OBJECTIVE: To investigate the effect of endogenous and exogenous oestrogen exposure on hearing levels in postmenopausal women. STUDY DESIGN: Retrospective cross-sectional study. SETTING: Population-based survey data collected by the Korean National Health and Nutrition Survey between 1 January 2010 and 31 December 2012. SUBJECTS AND METHODS: Participants comprised 3653 postmenopausal women. Detailed histories for reproductive factors and data on the use of hormone replacement therapy were obtained through health questionnaires and otologic examinations, including pure-tone audiogram and otoscopic findings. Complex-sample linear regression models controlling for confounding factors were generated to determine whether hormone-related factors were associated with hearing loss. RESULTS: Women who experienced a longer duration of oestrogen exposure had better hearing compared to those who do not in multivariate model adjusting for confounding factors with a lower adjusted beta coefficient of hearing threshold (ß = -0.18, 95% confidence interval = -0.3 to -0.07, P = .002). The results also suggested that hormone replacement therapy may be beneficial for attenuating hearing loss (ß = -1.22, 95% confidence interval = -2.19 to -0.25, P = .014), particularly in the high-frequency range from 3 to 6 KHz. CONCLUSION: A longer duration of lifetime oestrogen exposure (LEE) and the use of hormone replacement therapy are likely to attenuate hearing loss. These epidemiologic data provide evidence that oestrogen may be beneficial for attenuating age-related hearing decline.

Importance of High-Frequency Vestibular Function in the Prognosis of Bilateral Vestibulopathy.

OBJECTIVES: The aim of this study was to investigate whether preserved vestibular function in the high-frequency range influences the prognosis of patients with bilateral vestibulopathy (BVP) after vestibular rehabilitation. METHODS: Twenty-four patients followed up with vestibular rehabilitation were recruited. The enrolled patients were divided into two groups according to the preservation of the high-frequency vestibulo-ocular reflex (VOR) based on the video head impulse test (vHIT). The results of computerized dynamic posturography and the Dizziness Handicap Inventory (DHI) survey collected at baseline and at the 6-month follow-up after vestibular rehabilitation therapy were analyzed. RESULTS: Both groups showed significantly increased composite and DHI scores after follow-up with vestibular rehabilitation. The group with preserved high-frequency VOR showed a better composite score (p=0.064) and vestibular score (p= 0.008) than the group with lost high-frequency VOR at the 6-month follow up. The DHI score significantly decreased only in the group with lost high-frequency VOR (p=0.047). Among the three vestibular function tests (caloric test, rotary chair test, and vHIT) used to diagnose BVP, only vHIT showed a significant correlation (p=0.015) with a favorable prognosis (composite score ≥70). CONCLUSION: Better treatment outcomes are likely in patients with BVP with preserved vestibular function in response to high-frequency stimulation, as measured by the vHIT.

Feasibility of Asynchronous and Automated Telemedicine in Otolaryngology: Prospective Cross-Sectional Study.

BACKGROUND: COVID-19 often causes respiratory symptoms, making otolaryngology offices one of the most susceptible places for community transmission of the virus. Thus, telemedicine may benefit both patients and physicians. OBJECTIVE: This study aims to explore the feasibility of telemedicine for the diagnosis of all otologic disease types. METHODS: A total of 177 patients were prospectively enrolled, and the patient's clinical manifestations with otoendoscopic images were written in the electrical medical records. Asynchronous diagnoses were made for each patient to assess Top-1 and Top-2 accuracy, and we selected 20 cases to conduct a survey among four different otolaryngologists to assess the accuracy, interrater agreement, and diagnostic speed. We also constructed an experimental automated diagnosis system and assessed Top-1 accuracy and diagnostic speed. RESULTS: Asynchronous diagnosis showed Top-1 and Top-2 accuracies of 77.40% and 86.44%, respectively. In the selected 20 cases, the Top-2 accuracy of the four otolaryngologists was on average 91.25% (SD 7.50%), with an almost perfect agreement between them (Cohen kappa=0.91). The automated diagnostic model system showed 69.50% Top-1 accuracy. Otolaryngologists could diagnose an average of 1.55 (SD 0.48) patients per minute, while the machine learning model was capable of diagnosing on average 667.90 (SD 8.3) patients per minute. CONCLUSIONS: Asynchronous telemedicine in otology is feasible owing to the reasonable Top-2 accuracy when assessed by experienced otolaryngologists. Moreover, enhanced diagnostic speed while sustaining the accuracy shows the possibility of optimizing medical resources to provide expertise in areas short of physicians.

Machine learning approach for prediction of hearing preservation in vestibular schwannoma surgery.

In vestibular schwannoma patients with functional hearing status, surgical resection while preserving the hearing is feasible. Hearing levels, tumor size, and location of the tumor have been known to be candidates of predictors. We used a machine learning approach to predict hearing outcomes in vestibular schwannoma patients who underwent hearing preservation surgery: middle cranial fossa, or retrosigmoid approach. After reviewing the medical records of 52 patients with a pathologically confirmed vestibular schwannoma, we included 50 patient's records in the study. Hearing preservation was regarded as positive if the postoperative hearing was within serviceable hearing (50/50 rule). The categorical variable included the surgical approach, and the continuous variable covered audiometric and vestibular function tests, and the largest diameter of the tumor. Four different algorithms were lined up for comparison of accuracy: support vector machine(SVM), gradient boosting machine(GBM), deep neural network(DNN), and diffuse random forest(DRF). The average accuracy of predicting hearing preservation ranged from 62% (SVM) to 90% (DNN). The current study is the first to incorporate machine learning methodology into a prediction of successful hearing preservation surgery. Although a larger population may be needed for better generalization, this study could aid the surgeon's decision to perform a hearing preservation approach for vestibular schwannoma surgery.