A computational clinical decision-supporting system to suggest effective anti-epileptic drugs for pediatric epilepsy patients based on deep learning models using patient's medical history.

BACKGROUND: Epilepsy, a chronic brain disorder characterized by abnormal brain activity that causes seizures and other symptoms, is typically treated using anti-epileptic drugs (AEDs) as the first-line therapy. However, due to the variations in their modes of action, identification of effective AEDs often relies on ad hoc trials, which is particularly challenging for pediatric patients. Thus, there is significant value in computational methods capable of assisting in the selection of AEDs, aiming to minimize unnecessary medication and improve treatment efficacy. RESULTS: In this study, we collected 7,507 medical records from 1,000 pediatric epilepsy patients and developed a computational clinical decision-supporting system for AED selection. This system leverages three multi-channel convolutional neural network (CNN) models tailored to three specific AEDs (vigabatrin, prednisolone, and clobazam). Each CNN model predicts whether a respective AED is effective on a given patient or not. The CNN models showed AUROCs of 0.90, 0.80, and 0.92 in 10-fold cross-validation, respectively. Evaluation on a hold-out test dataset further revealed positive predictive values (PPVs) of 0.92, 0.97, and 0.91 for the three respective CNN models, representing that suggested AEDs by our models would be effective in controlling epilepsy with a high accuracy and thereby reducing unnecessary medications for pediatric patients. CONCLUSION: Our CNN models in the system demonstrated high PPVs for the three AEDs, which signifies the potential of our approach to support the clinical decision-making by assisting doctors in recommending effective AEDs within the three AEDs for patients based on their medical history. This would result in a reduction in the number of unnecessary ad hoc attempts to find an effective AED for pediatric epilepsy patients.

Diagnostic accuracy of clinical visual assessment using endoscopic images for nasal cavity mass lesions.

Objectives: Discrimination of nasal cavity lesions using nasal endoscopy is challenging because of the differences in clinical manifestations and treatment strategies. We aimed to investigate the diagnostic accuracy of clinical visual assessment (CVA) of nasal cavity masses using endoscopic images and determine whether there is a difference according to pathologic class and the examiners' experience. Methods: We collected pathologically confirmed endoscopic images of normal findings, nasal polyp (NP), benign tumor, and malignant tumor (each class contained 100 images) randomly selected. Eighteen otolaryngologists, including six junior residents, six senior residents, and six board-certified rhinologists classified the test set images into four classes of lesions by CVA. Diagnostic performance according to the pathologic class and the examiner's experience level was evaluated based on overall accuracy, F1-score, confusion matrix, and area under the receiver operating characteristic curve (AUC). Results: Diagnostic performance was significantly different according to the pathological class of nasal cavity mass lesions with the overall accuracy reported high in the order of normal, NP, benign tumor, and malignant tumor (0.926 ± 0.100; 0.819 ± 0.135; 0.580 ± 0.112; 0.478 ± 0.187, respectively), F1 score (0.937 ± 0.076; 0.730 ± 0.093; 0.549 ± 0.080; 0.554 ± 0.146, respectively) and AUC value (0.96 ± 0.06; 0.84 ± 0.07; 0.70 ± 0.05; 0.71 ± 0.08, respectively). The expert rhinologist group achieved higher overall accuracy than the resident group (0.756 ± 0.157 vs. 0.680 ± 0.239, p < .05). Conclusion: CVA for nasal cavity mass was highly dependent on the pathologic class and examiner's experience. The overall accuracy was reliably high for normal findings, but low in classifying benign and malignant tumors. Differential diagnosis of lesions solely based on nasal endoscopic evaluation is challenging. Therefore, clinicians should consider further clinical evaluation for suspicious cases.

Deep learning algorithm for the automated detection and classification of nasal cavity mass in nasal endoscopic images.

Nasal endoscopy is routinely performed to distinguish the pathological types of masses. There is a lack of studies on deep learning algorithms for discriminating a wide range of endoscopic nasal cavity mass lesions. Therefore, we aimed to develop an endoscopic-examination-based deep learning model to detect and classify nasal cavity mass lesions, including nasal polyps (NPs), benign tumors, and malignant tumors. The clinical feasibility of the model was evaluated by comparing the results to those of manual assessment. Biopsy-confirmed nasal endoscopic images were obtained from 17 hospitals in South Korea. Here, 400 images were used for the test set. The training and validation datasets consisted of 149,043 normal nasal cavity, 311,043 NP, 9,271 benign tumor, and 5,323 malignant tumor lesion images. The proposed Xception architecture achieved an overall accuracy of 0.792 with the following class accuracies on the test set: normal = 0.978 ± 0.016, NP = 0.790 ± 0.016, benign = 0.708 ± 0.100, and malignant = 0.698 ± 0.116. With an average area under the receiver operating characteristic curve (AUC) of 0.947, the AUC values and F1 score were highest in the order of normal, NP, malignant tumor, and benign tumor classes. The classification performances of the proposed model were comparable with those of manual assessment in the normal and NP classes. The proposed model outperformed manual assessment in the benign and malignant tumor classes (sensitivities of 0.708 ± 0.100 vs. 0.549 ± 0.172, 0.698 ± 0.116 vs. 0.518 ± 0.153, respectively). In urgent (malignant) versus nonurgent binary predictions, the deep learning model achieved superior diagnostic accuracy. The developed model based on endoscopic images achieved satisfactory performance in classifying four classes of nasal cavity mass lesions, namely normal, NP, benign tumor, and malignant tumor. The developed model can therefore be used to screen nasal cavity lesions accurately and rapidly.

Corydalis ternata Nakai Alleviates Cognitive Decline in Alzheimer's Disease by Reducing ß-Amyloid and Neuroinflammation.

Recently, natural herbs have gained increasing attention owing to their comparatively low toxicity levels and the abundance of historical medical documentation regarding their use. Nevertheless, owing to a lack of knowledge regarding these herbs and their compounds, attempts to find those that could be beneficial for treating diseases have often been ad hoc; thus, there is now a growing demand for an in silico method to identify beneficial herbs. In this study, we present a computational approach for identifying natural herbs specifically effective in treating cognitive decline in Alzheimer's disease (AD) sufferers, which analyzes the similarities between herbal compounds and known drugs targeting AD-related proteins. Our in silico method suggests that Corydalis ternata can improve cognitive decline in AD sufferers. Behavioral tests with an AD mouse model for the confirmation of the in silico prediction reveals that C. ternata significantly alleviated the cognitive decline (memory and motor functions) caused by neurodegeneration. Further pathology analyses reveal that C. ternata decreases the level of Aß plaques, reduces neuroinflammation, and promotes autophagy flux, and thus C. ternata can be clinically effective for preventing mild cognitive impairment during the early stages of AD. These findings highlight the potential utility of our in silico method and the potential clinical application of the identified natural herb in treating and preventing AD.

Impact of long-term nasal airflow deprivation on sinonasal structures and chronic rhinosinusitis in total laryngectomy patients.

Objective: Total laryngectomy (TL) patients are good models in which to evaluate the effects of nasal airflow cessation on the sinonasal tract. Here, we evaluated changes in sinonasal structures and association with sinus opacification in the computed tomography (CT) images 3 years post-TL. Methods: Patients who underwent TL from 2005 to 2017 in a teaching academic center were reviewed retrospectively. Patients with a final follow-up CT taken less than 3 years after TL, tracheoesophageal puncture, inadequate CT image, or history of sinonasal surgery were excluded. The control group included patients who underwent a partial laryngectomy or hypopharyngectomy without requiring a tracheotomy for more than a month. Altogether, 45 TL patients and 38 controls were selected. The volume of all four paranasal sinuses, inferior turbinate soft tissue volume (ITSTV), maxillary sinus natural ostium (MSNO) mucosal width, and Lund-Mackay scores (LMS) were measured on preoperative and postoperative CT scans. Results: The mean duration between surgery and the final CT scan was 6.3 ± 2.4 and 5.5 ± 2.3 years for the TL and control groups, respectively. Neither group showed significant changes in the four paranasal sinuses' volume or MSNO mucosa width. The ITSTV decreased significantly, from 4.6 ± 1.3 to 2.8 ± 1.1 mL (p < .001), in the TL group, regardless of the presence of nasal septal deviation, showing ITSTV reduction on both concave and convex sides. By contrast, the control group showed no significant changes in ITSTV. Postoperative LMS changes in both groups were insignificant. The number of patients with LMS aggravation or alleviation was the same in both groups, regardless of preoperative sinus opacification. Conclusions: Paranasal sinus structures and sinus opacification are not affected significantly by nasal airflow cessation; however, the inferior turbinate mucosa is affected by long-term discontinuation of nasal airflow. Level of Evidence: 4 (case-control study).

Sinonasal Organizing Hematoma: Demographics, Diagnosis, and Treatment Outcomes of 112 patients.

INTRODUCTION: Sinonasal organizing hematoma (OH) is a rare, nonneoplastic lesion that often presents with epistaxis, a reddish mass, and destruction of the involved sinonasal structures. Due to its rarity, the demographics, diagnostic modalities, treatment strategies, and outcomes have not yet been studied in a large, long-term study. MATERIALS AND METHODS: Retrospect cohort of 112 sinonasal OH patients treated between 1997 and 2020 in a tertiary, university hospital were evaluated. Demographics, systemic comorbidities, sinonasal surgery history, serum laboratory tests, radiological findings, and treatment results were collected. The present study aimed to assess the accuracy of preoperative computed tomography (CT), Gadolinum-enhanced magnetic resonance (MR), and punch biopsies in detecting sinonasal OH as the most likely diagnosis. In addition, incidental differences by age and year of diagnosis were calculated using the Poisson log-linear regression model. RESULTS: The median age was 44, and 58% were male. Fewer than 20% of these cases had a chronic systemic comorbidity, bleeding tendency, or sinonasal surgery history. MR had the highest accuracy of (87%) to detect sinonasal OH as the most likely diagnosis, compared with contrast-enhanced-CT (53%), punch biopsy (49%), and non-enhanced-CT (16%) (all <0.05). Sinonasal OH incidence did not vary by age, but the yearly rate significantly increased by 1.05 times over 23 years (p < 0.05). Notably, 84% of 112 patients received surgical removal through the assistance of an endoscope, and none had substantial bleeding without preoperative embolization. CONCLUSION: Sinonasal OH was observed regardless of age, sex, systemic comorbidities, bleeding tendency, prior sinonasal surgery, or trauma. Preoperative MR gives the highest accuracy for detecting this disease. Sinonasal OH may be safely managed with endoscopic-assisted surgery removal without embolization. LEVEL OF EVIDENCE: 4 Laryngoscope, 134:1581-1590, 2024.

A multi-layered network model identifies Akt1 as a common modulator of neurodegeneration.

The accumulation of misfolded and aggregated proteins is a hallmark of neurodegenerative proteinopathies. Although multiple genetic loci have been associated with specific neurodegenerative diseases (NDs), molecular mechanisms that may have a broader relevance for most or all proteinopathies remain poorly resolved. In this study, we developed a multi-layered network expansion (MLnet) model to predict protein modifiers that are common to a group of diseases and, therefore, may have broader pathophysiological relevance for that group. When applied to the four NDs Alzheimer's disease (AD), Huntington's disease, and spinocerebellar ataxia types 1 and 3, we predicted multiple members of the insulin pathway, including PDK1, Akt1, InR, and sgg (GSK-3ß), as common modifiers. We validated these modifiers with the help of four Drosophila ND models. Further evaluation of Akt1 in human cell-based ND models revealed that activation of Akt1 signaling by the small molecule SC79 increased cell viability in all models. Moreover, treatment of AD model mice with SC79 enhanced their long-term memory and ameliorated dysregulated anxiety levels, which are commonly affected in AD patients. These findings validate MLnet as a valuable tool to uncover molecular pathways and proteins involved in the pathophysiology of entire disease groups and identify potential therapeutic targets that have relevance across disease boundaries. MLnet can be used for any group of diseases and is available as a web tool at http://ssbio.cau.ac.kr/software/mlnet.

A machine learning-based quantitative model (LogBB_Pred) to predict the blood-brain barrier permeability (logBB value) of drug compounds.

MOTIVATION: Efficient assessment of the blood-brain barrier (BBB) penetration ability of a drug compound is one of the major hurdles in central nervous system drug discovery since experimental methods are costly and time-consuming. To advance and elevate the success rate of neurotherapeutic drug discovery, it is essential to develop an accurate computational quantitative model to determine the absolute logBB value (a logarithmic ratio of the concentration of a drug in the brain to its concentration in the blood) of a drug candidate. RESULTS: Here, we developed a quantitative model (LogBB_Pred) capable of predicting a logBB value of a query compound. The model achieved an R2 of 0.61 on an independent test dataset and outperformed other publicly available quantitative models. When compared with the available qualitative (classification) models that only classified whether a compound is BBB-permeable or not, our model achieved the same accuracy (0.85) with the best qualitative model and far-outperformed other qualitative models (accuracies between 0.64 and 0.70). For further evaluation, our model, quantitative models, and the qualitative models were evaluated on a real-world central nervous system drug screening library. Our model showed an accuracy of 0.97 while the other models showed an accuracy in the range of 0.29-0.83. Consequently, our model can accurately classify BBB-permeable compounds as well as predict the absolute logBB values of drug candidates. AVAILABILITY AND IMPLEMENTATION: Web server is freely available on the web at http://ssbio.cau.ac.kr/software/logbb_pred/. The data used in this study are available to download at http://ssbio.cau.ac.kr/software/logbb_pred/dataset.zip.

Preoperative Prediction of Sinonasal Inverted Papilloma-associated Squamous Cell Carcinoma (IP-SCC).

INTRODUCTION: Sinonasal inverted papillomas (IP) can undergo transformation into IP-squamous cell carcinomas (IP-SCC). More aggressive treatment plan should be established when IP-SCC is suspected. Nevertheless, inaccuracy of the preoperative punch biopsy results to detect IP-SCC from IP raises the need for an additional strategy. The present study aimed to investigate significant clinicoradiological remarks associated with IP-SCC than IP. MATERIAL AND METHODS: Postoperative surgical specimens obtained from patients with confirmed IP or IP-SCC at a single tertiary medical center from 1997 to 2018 were retrospectively evaluated. Patients' demographic and clinical characteristics, preoperative in-office punch biopsy results, and preoperative computed tomography (CT) or magnetic resonance images were reviewed. Univariate and multivariate analyses were performed to assess the odds ratio (OR) associated with IP-SCC. The area under the curve (AUC) in the receiver Operating Characteristic (ROC) curve was calculated in the prediction model to discriminate IP-SCC from IP. RESULTS: The study included 44 IP-SCC and 301 patients with IP. The diagnostic sensitivity of in-office punch biopsy to detect IP-SCC was 70.7%. Multivariate analysis showed that factors significantly associated with IP-SCC included tobacco smoking >10PY (adjusted-OR [aOR]: 4.1), epistaxis (aOR: 3.4), facial pain (aOR: 4.2), bony destruction (aOR: 37.6), bony remodeling (aOR: 36.3), and invasion of adjacent structures (aOR: 31.6) (all p < 0.05). Combining all significantly related clinicoradiological features, the ability to discriminate IP-SCC from IP reached an AUC of 0.974. CONCLUSION: IP patients with a history of tobacco smoking, facial pain, epistaxis, and bony destruction, remodeling, or invasion of an adjacent structure on preoperative images may be at higher risk for IP-SCC. LEVEL OF EVIDENCE: 3 Laryngoscope, 133:2502-2510, 2023.

Features of Odontogenic Sinusitis Associated With Dental Implants.

OBJECTIVES: With the increase in dental implants for tooth loss, odontogenic sinusitis following maxillary dental implants is frequently encountered in otorhinolaryngology practice. The authors aimed to reveal the association between implant extrusion into maxillary sinus, along with implant-related complications in patients diagnosed with implant-related odontogenic sinusitis (IR-ODS). STUDY DESIGN: Case-control study. METHODS: This study enrolled 60 patients who received functional endoscopic sinus surgery due to IR-ODS. The preoperative sinus computed tomography was retrospectively reviewed. Among the 120 maxillary sinuses of the 60 patients, 68 sides were diagnosed with IR-ODS sides, whereas 27 sides showed no clinical or radiological evidence of this condition after the implant insertion and were defined as the control sides. Statistical analysis between these two groups was conducted, in addition to odds ratio (OR) calculations for associations with IR-ODS. RESULTS: The mean age of the IR-ODS subjects was 59.5 ± 19.1, with a male to female ratio of 32/28 (53.3%/46.7%). Implants extruding by more than 4 mm into the maxillary sinus, peri-implantitis, bone graft disruption-extrusion were associated with a significantly higher incidence in the IR-ODS (p = 0.035, p = 0.003, p = 0.011, respectively). The IR-ODS sides showed an adjusted-OR (95% confidence interval) of 27.4 (2.7-276.5) for extrusion length >4 mm, 11.8 (3.0-46.5) for peri-implantitis, and 34.1 (3.3-347.8) for bone graft disruption (p = 0.005, p < 0.001, and p = 0.003, respectively). CONCLUSION: Maxillary dental implants extruding more than 4 mm into the maxillary sinus, peri-implantitis, and disrupted-extruded bone grafts show significant association with IR-ODS. LEVEL OF EVIDENCE: 4 Laryngoscope, 133:237-243, 2023.

Analysis methods for the gut microbiome in neuropsychiatric and neurodegenerative disorders.

For a long time, the central nervous system was believed to be the only regulator of cognitive functions. However, accumulating evidence suggests that the composition of the microbiome is strongly associated with brain functions and diseases. Indeed, the gut microbiome is involved in neuropsychiatric diseases (e.g., depression, autism spectrum disorder, and anxiety) and neurodegenerative diseases (e.g., Parkinson's disease and Alzheimer's disease). In this review, we provide an overview of the link between the gut microbiome and neuropsychiatric or neurodegenerative disorders. We also introduce analytical methods used to assess the connection between the gut microbiome and the brain. The limitations of the methods used at present are also discussed. The accurate translation of the microbiome information to brain disorder could promote better understanding of neuronal diseases and aid in finding alternative and novel therapies.

Feasibility of distraction osteogenesis maxillary expansion in patients with persistent nasal obstruction after septoplasty.

BACKGROUND: Distraction osteogenesis maxillary expansion (DOME) is an emerging surgical concept for obstructive sleep apnea patients with a high-arched and narrow hard palate presenting with nasal obstruction. Herein we demonstrate the feasibility of DOME in the treatment of patients with persistent nasal obstruction after primary septoplasty. METHODS: A total of 32 consecutive patients with persistent nasal obstruction after septoplasty who underwent DOME were enrolled in this study. The efficacy of DOME in treating persistent nasal obstruction in septoplasty failure patients was evaluated using the validated Nasal Obstruction Symptom Evaluation (NOSE) scale. Pre- and postoperative computed tomography (CT) scans were performed to measure the change of nasal floor width after surgery. RESULTS: NOSE score improved significantly from 13.62 ± 4.74 to 5.15 ± 3.50 (p < 0.001) after DOME. The postoperative CT scan showed a significant increase of the nasal floor width from 22.4 ± 4.21 to 28.06 ± 4.52 (p < 0.01). No patient had major complications such as wound infection, oronasal fistula, sinusitis, loss of incisor teeth, or malunion. CONCLUSION: We found that DOME appears to improve persistent nasal obstruction after septoplasty in patients with a narrow and high-arched hard palate.

Construction of a tunable promoter library to optimize gene expression in Methylomonas sp. DH-1, a methanotroph, and its application to cadaverine production.

BACKGROUND: As methane is 84 times more potent than carbon dioxide in exacerbating the greenhouse effect, there is an increasing interest in the utilization of methanotrophic bacteria that can convert harmful methane into various value-added compounds. A recently isolated methanotroph, Methylomonas sp. DH-1, is a promising biofactory platform because of its relatively fast growth. However, the lack of genetic engineering tools hampers its wide use in the bioindustry. RESULTS: Through three different approaches, we constructed a tunable promoter library comprising 33 promoters that can be used for the metabolic engineering of Methylomonas sp. DH-1. The library had an expression level of 0.24-410% when compared with the strength of the lac promoter. For practical application of the promoter library, we fine-tuned the expressions of cadA and cadB genes, required for cadaverine synthesis and export, respectively. The strain with PrpmB-cadA and PDnaA-cadB produced the highest cadaverine titre (18.12 ± 1.06 mg/L) in Methylomonas sp. DH-1, which was up to 2.8-fold higher than that obtained from a non-optimized strain. In addition, cell growth and lysine (a precursor of cadaverine) production assays suggested that gene expression optimization through transcription tuning can afford a balance between the growth and precursor supply. CONCLUSIONS: The tunable promoter library provides standard and tunable components for gene expression, thereby facilitating the use of methanotrophs, specifically Methylomonas sp. DH-1, as a sustainable cell factory.

A Novel Vitronectin Peptide Facilitates Differentiation of Oligodendrocytes from Human Pluripotent Stem Cells (Synthetic ECM for Oligodendrocyte Differentiation).

Differentiation of oligodendrocytes (ODs) presents a challenge in regenerative medicine due to their role in various neurological diseases associated with dysmyelination and demyelination. Here, we designed a peptide derived from vitronectin (VN) using in silico docking simulation and examined its use as a synthetic substrate to support the differentiation of ODs derived from human pluripotent stem cells. The designed peptide, named VNP2, promoted OD differentiation induced by the overexpression of SOX10 in OD precursor cells compared with Matrigel and full-length VN. ODs differentiated on VNP2 exhibited greater contact with axon-mimicking nanofibers than those differentiated on Matrigel. Transcriptomic analysis revealed that the genes associated with morphogenesis, cytoskeleton remodeling, and OD differentiation were upregulated in cells grown on VNP2 compared with cells grown on Matrigel. This new synthetic VN-derived peptide can be used to develop a culture environment for efficient OD differentiation.

Feasibility of a deep learning-based algorithm for automated detection and classification of nasal polyps and inverted papillomas on nasal endoscopic images.

BACKGROUND: Discrimination of nasal cavity mass lesions is a challenging work requiring extensive experience. A deep learning-based automated diagnostic system may help clinicians to classify nasal cavity mass lesions. We demonstrated the feasibility of a convolutional neural network (CNN)-based diagnosis system for automatic detection and classification of nasal polyps (NP) and inverted papillomas (IP). METHODS: We developed a CNN-based algorithm using a transfer learning strategy and trained it on nasal endoscopic images. A total of 99 nasal endoscopic images with normal findings, 98 images with NP, and 100 images with IP were analyzed using the developed CNN. Six otolaryngologists participated in clinical visual assessment. Image-based classification performance was measured by calculating the accuracy and area under the receiver operating characteristic curve (AUC). The diagnostic performance was compared between the CNN and clinical visual assessment by human experts. RESULTS: The algorithm achieved an overall accuracy of 0.742 ± 0.058 with the following class accuracies: normal, 0.81± 0.14; IP, 0.57 ± 0.07; and NP, 0.83 ± 0.21. The AUC values for normal, IP, and NP were 0.91 ± 0.06, 0.82 ± 0.09, and 0.84 ± 0.06, respectively. The overall accuracy of the CNN model was comparable with the average performance of human experts (0.742 vs. 0.749; p = 0.11). CONCLUSIONS: The trained CNN model appears to reliably classify NP and IP of the nasal cavity from nasal endoscopic images; it also yields a reliable reference for diagnosing nasal cavity mass lesions during nasal endoscopy. However, further studies with more test data are warranted to improve the diagnostic accuracy of our CNN model.

Adenotonsillectomy Does not Alter the Risk of Upper Airway Infections in Children.

OBJECTIVE: Given the conflicting evidence regarding adenotonsillectomy and the risk of upper airway infections (UAIs), including acute pharyngitis (APT), rhinosinusitis (ARS), and otitis media (AOM), we evaluated the risk of developing UAIs following adenotonsillectomy in a childhood population. METHODS: In this population-based follow-up study, we used data from the National Health Insurance Service-National Sample Cohort. The adenotonsillectomy group (n = 2,377) included patients aged <10 years who underwent an adenotonsillectomy. For every patient with APT (n = 2,309), ARS (n = 2,308), and AOM (n = 2,207) who had an adenotonsillectomy, four participants were randomly selected for the control groups (n = 9,204, n = 9,196, and n = 8,788, respectively) using propensity score matching. The number of postoperative hospital visits for UAIs was recorded for 1 to 9 years, and the equivalence test was used to compare the number of visits between the adenotonsillectomy and control groups. RESULTS: There were no significant differences in the incidence of APT, ARS, or AOM at the margin of equivalence of difference (-0.5 < 95% confidence interval of difference < 0.5) between the two groups from postoperative year 1 to year 9. The hospital visits for these diseases gradually decreased over time in both groups. There was a significant decrease in the number of visits for APT, ARS, and AOM in the 5-9 years age group when compared with those in the <4 years age group (P < .01). CONCLUSION: Our results suggest that adenotonsillectomy does not alter the frequency of UAIs in children. LEVEL OF EVIDENCE: NA Laryngoscope, 131:2376-2383, 2021.

LightBBB: computational prediction model of blood-brain-barrier penetration based on LightGBM.

MOTIVATION: Identification of blood-brain barrier (BBB) permeability of a compound is a major challenge in neurotherapeutic drug discovery. Conventional approaches for BBB permeability measurement are expensive, time-consuming and labor-intensive. BBB permeability is associated with diverse chemical properties of compounds. However, BBB permeability prediction models have been developed using small datasets and limited features, which are usually not practical due to their low coverage of chemical diversity of compounds. Aim of this study is to develop a BBB permeability prediction model using a large dataset for practical applications. This model can be used for facilitated compound screening in the early stage of brain drug discovery. RESULTS: A dataset of 7162 compounds with BBB permeability (5453 BBB+ and 1709 BBB-) was compiled from the literature, where BBB+ and BBB- denote BBB-permeable and non-permeable compounds, respectively. We trained a machine learning model based on Light Gradient Boosting Machine (LightGBM) algorithm and achieved an overall accuracy of 89%, an area under the curve (AUC) of 0.93, specificity of 0.77 and sensitivity of 0.93, when 10-fold cross-validation was performed. The model was further evaluated using 74 central nerve system compounds (39 BBB+ and 35 BBB-) obtained from the literature and showed an accuracy of 90%, sensitivity of 0.85 and specificity of 0.94. Our model outperforms over existing BBB permeability prediction models. AVAILABILITYAND IMPLEMENTATION: The prediction server is available at http://ssbio.cau.ac.kr/software/bbb.

2-D chemical structure image-based in silico model to predict agonist activity for androgen receptor.

BACKGROUND: Abnormal activation of human nuclear hormone receptors disrupts endocrine systems and thereby affects human health. There have been machine learning-based models to predict androgen receptor agonist activity. However, the models were constructed based on limited numerical features such as molecular descriptors and fingerprints. RESULT: In this study, instead of the numerical features, 2-D chemical structure images of compounds were used to build an androgen receptor toxicity prediction model. The images may provide unknown features that were not represented by conventional numerical features. As a result, the new strategy resulted in a construction of highly accurate prediction model: Mathews correlation coefficient (MCC) of 0.688, positive predictive value (PPV) of 0.933, sensitivity of 0.519, specificity of 0.998, and overall accuracy of 0.981 in 10-fold cross-validation. Validation on a test dataset showed MCC of 0.370, sensitivity of 0.211, specificity of 0.991, PPV of 0.882, and overall accuracy of 0.801. Our chemical image-based prediction model outperforms conventional models based on numerical features. CONCLUSION: Our constructed prediction model successfully classified molecular images into androgen receptor agonists or inactive compounds. The result indicates that 2-D molecular mimetic diagram would be used as another feature to construct molecular activity prediction models.

Maxillomandibular Advancement and Upper Airway Stimulation: Extrapharyngeal Surgery for Obstructive Sleep Apnea.

There are many ways to categorize surgery for obstructive sleep apnea (OSA), one of which is to distinguish between intrapharyngeal and extrapharyngeal procedures. While the general otolaryngologist treating OSA is familiar with intrapharyngeal procedures, such as uvulopalatopharyngoplasty and tongue base reduction, extrapharyngeal sleep operations such as maxillomandibular advancement (MMA) and upper airway stimulation (UAS) have evolved rapidly in the recent decade and deserve a dedicated review. MMA and UAS have both shown predictable high success rates with low morbidity. Each approach has unique strengths and limitations, and for the most complex of OSA patients, the two in combination complement each other. Extrapharyngeal airway operations are critical for achieving favorable outcomes for sleep surgeons.