Abstract Background Both predictions and predictors of difficult laryngoscopy are controversial. Machine learning is an excellent alternative method for predicting difficult laryngoscopy. This study aimed to develop and validate practical predictive models for difficult laryngoscopy through machine learning. Methods Variables for the prediction of difficult laryngoscopy included age, Mallampati grade, body mass index, sternomental distance, and neck circumference. Difficult laryngoscopy was defined as grade 3 and 4 by the Cormack-Lehane classification. Pre-anesthesia and anesthesia data of 616 patients who had undergone anesthesia at a single center were included. The dataset was divided into a base training set (n = 492) and a base test set (n = 124), with equal distribution of difficult laryngoscopy. Training data sets were trained with six algorithms (multilayer perceptron, logistic regression, supportive vector machine, random forest, extreme gradient boosting, and light gradient boosting machine), and cross-validated. The model with the highest area under the receiver operating characteristic curve (AUROC) was chosen as the final model, which was validated with the test set. Results The results of cross-validation were best using the light gradient boosting machine algorithm with Mallampati score x age and sternomental distance as predictive model parameters. The predicted AUROC for the difficult laryngoscopy class was 0.71 (95% confidence interval, 0.59-0.83; p= 0.014), and the recall (sensitivity) was 0.85. Conclusion Predicting difficult laryngoscopy is possible with three parameters. Severe damage resulting from failure to predict difficult laryngoscopy with high recall is small with the reported model. The model’s performance can be further enhanced by additional data training.

Purpose The study evaluated whether preoperative measures of the C-reactive protein-based systemic inflammatory response may predict cancer survival independent of tumor stage in patients with upper urinary tract urothelial carcinoma (UTUC). Materials and Methods Between September 1999 and October 2010, 181 patients submitted to radical nephroureterectomy were available for evaluation. Multivariate survival analyses were performed using Cox’s proportional hazards model and the coefficient for each factor was divided by the highest coefficient, multiplied by 4, and rounded to the nearest integer. Results Multivariate analyses showed that tumor location, pathologic T stage, lymphovascular invasion, margin status, and albumin level were independent contributors. The bootstrap-corrected C statistics of the model were 0.813 for disease-specific survival and 0.755 for overall survival, respectively. For time to disease-specific and overall mortality for patients, integrated area under the curve values were 0.792 and 0.739, respectively. When patients were clustered into three groups according to their model-predicted survival, the 5-year disease-specific survival in the low-, intermediate- and high-risk group was 95.4%, 76.2%, and 36.9%, respectively (p<0.001), and were 87.8%, 54.4%, and 31.8%, respectively, for overall survival (p<0.001). Decision curve analysis revealed that the use of model was associated with net benefit gains relative to the treat-all strategy. Conclusions   Pretreatment albumin is a simple biomarker based on routinely available well-standardized measures, and is not an expensive and time-consuming process. Hypoalbuminemia is an independent marker of poor prognosis in patients with upper urinary tract urothelial carcinoma.