Prognostic value of the combination of neutrophil-to-lymphocyte ratio, monocyte-to-lymphocyte ratio and platelet-to-lymphocyte ratio on mortality in patients on maintenance hemodialysis.

BACKGROUND: Hemodialysis (HD) is the most important renal replacement therapy for patients with end-stage kidney disease (ESKD). Systemic inflammation is a risk factor of mortality in HD patients. Neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), and platelet-to-lymphocyte ratio (PLR) are new inflammatory markers. However, previous studies have inconsistent conclusions about the predictive value of NLR, MLR and PLR on mortality of HD patients. The aim of this study was to establish an inflammation scoring system by including NLR, MLR and PLR, and evaluate the association between the inflammation score and all-cause and cardiovascular mortality in HD patients. METHODS: In this single center retrospective cohort study, 213 incident HD patients from January 1, 2015 to December 31, 2020 were included. Baseline demographic and clinical data and laboratory measurements were collected. According to the optimal cut-off values, NLR, MLR and PLR were assigned 0 or 1 point, respectively. Then, the inflammation score was obtained by adding the NLR, MLR and PLR scores. All patients were followed until July 31, 2021. The associations of the inflammation score with all-cause and cardiovascular mortality were assessed by multivariable-adjusted Cox models. RESULTS: Of 213 patients, the mean (± SD) age was 56.8 ± 14.4 years, 66.2% were men, and 32.9% with diabetes. The primary cause of ESKD was mainly chronic glomerulonephritis (46.5%) and diabetic nephropathy (28.6%). The median inflammation score was 2 (interquartile range = 1-3). During a median 30 months (interquartile range = 17-50 months) follow-up period, 53 patients had died, of which 33 deaths were caused by cardiovascular disease. After adjusting for demographics, primary diseases and other confounders in multivariable model, the inflammation score = 3 was associated with a hazard ratio for all-cause mortality of 4.562 (95% confidence interval, 1.342-15.504, P = 0.015) and a hazard ratio for cardiovascular mortality of 4.027 (95% confidence interval, 0.882-18.384, P = 0.072). CONCLUSION: In conclusion, an inflammation scoring system was established by including NLR, MLR and PLR, and the higher inflammation score was independently associated with all-cause mortality in HD patients.

Feature Fusion of a Deep-Learning Algorithm into Wearable Sensor Devices for Human Activity Recognition.

This paper presents a wearable device, fitted on the waist of a participant that recognizes six activities of daily living (walking, walking upstairs, walking downstairs, sitting, standing, and laying) through a deep-learning algorithm, human activity recognition (HAR). The wearable device comprises a single-board computer (SBC) and six-axis sensors. The deep-learning algorithm employs three parallel convolutional neural networks for local feature extraction and for subsequent concatenation to establish feature fusion models of varying kernel size. By using kernels of different sizes, relevant local features of varying lengths were identified, thereby increasing the accuracy of human activity recognition. Regarding experimental data, the database of University of California, Irvine (UCI) and self-recorded data were used separately. The self-recorded data were obtained by having 21 participants wear the device on their waist and perform six common activities in the laboratory. These data were used to verify the proposed deep-learning algorithm on the performance of the wearable device. The accuracy of these six activities in the UCI dataset and in the self-recorded data were 97.49% and 96.27%, respectively. The accuracies in tenfold cross-validation were 99.56% and 97.46%, respectively. The experimental results have successfully verified the proposed convolutional neural network (CNN) architecture, which can be used in rehabilitation assessment for people unable to exercise vigorously.