Indoor Localization Algorithm Based on a High-Order Graph Neural Network.

Given that fingerprint localization methods can be effectively modeled as supervised learning problems, machine learning has been employed for indoor localization tasks based on fingerprint methods. However, it is often challenging for popular machine learning models to effectively capture the unstructured data features inherent in fingerprint data that are generated in diverse propagation environments. In this paper, we propose an indoor localization algorithm based on a high-order graph neural network (HoGNNLoc) to enhance the accuracy of indoor localization and improve localization stability in dynamic environments. The algorithm first designs an adjacency matrix based on the spatial relative locations of access points (APs) to obtain a graph structure; on this basis, a high-order graph neural network is constructed to extract and aggregate the features; finally, the designed fully connected network is used to achieve the regression prediction of the location of the target to be located. The experimental results on our self-built dataset show that the proposed algorithm achieves localization accuracy within 1.29 m at 80% of the cumulative distribution function (CDF) points. The improvements are 59.2%, 51.3%, 36.1%, and 22.7% compared to the K-nearest neighbors (KNN), deep neural network (DNN), simple graph convolutional network (SGC), and graph attention network (GAT). Moreover, even with a 30% reduction in fingerprint data, the proposed algorithm exhibits stable localization performance. On a public dataset, our proposed localization algorithm can also show better performance.

Inflammation-related nomogram for predicting survival of patients with unresectable hepatocellular carcinoma received conversion therapy.

BACKGROUND: The efficacy of conversion therapy for patients with unresectable hepatocellular carcinoma (HCC) is a common clinical concern. AIM: To analyse the prognostic factors of overall survival (OS) in patients with unresectable HCC who received conversion therapy. METHODS: One hundred and fifty patients who met the inclusion criteria were enrolled and divided into a training cohort (n = 120) and a validation cohort (n = 30). Using the independent risk factors in the training cohort, a nomogram model was constructed to predict OS for patients treated with transarterial chemoembolization following hepatic resection. The nomogram was internally validated with the bootstrapping method. The predictive performance of nomogram was assessed by Harrell's concordance index (C-index), calibration plot and time-dependent receiver operating characteristic curves and compared with six other conventional HCC staging systems. RESULTS: Multivariate Cox analysis identified that albumin, blood urea nitrogen, gamma-glutamyl transpeptidase to platelet ratio, platelet to lymphocyte ratio, macrovascular invasion and tumour number were the six independent prognostic factors correlated with OS in nomogram model. The C-index in the training cohort and validation cohort were 0.752 and 0.807 for predicting OS, which were higher than those of the six conventional HCC staging systems (0.563 to 0.715 for the training cohort and 0.458 to 0.571 for the validation cohort). The calibration plots showed good consistency between the nomogram prediction of OS and the actual observations of OS. Decision curve analyses indicated satisfactory clinical utility. With a total nomogram score of 196, patients were accurately classified into low-risk and high-risk groups. Furthermore, we have deployed the model into online calculators that can be accessed for free at https://ctmodelforunresectablehcc.shinyapps.io/DynNomapp/. CONCLUSION: The nomogram achieved optimal individualized prognostication of OS in HCC patients who received conversion therapy, which could be a useful clinical tool to help guide postoperative personalized interventions and prognosis judgement.

Microfluidic paper-based aptasensor devices for multiplexed detection of pathogenic bacteria.

Foodborne pathogens are major public health concerns worldwide. Paper-based microfluidic devices are versatile, user friendly and low cost. We report a novel paper-based single input channel microfluidic device that can detect more than one whole-cell foodborne bacteria at the same time, and comes with quantitative reading via image analysis. This microfluidic paper-based multiplexed aptasensor simultaneously detects E. coli O157:H7 and S. Typhimurium. Custom designed particles provide colorimetric signal enhancement and false results prevention. Several aptamers were screened and the highest-affinity aptamers were optimized and employed for detection of these bacteria in solution, both in a buffer as well as pear juice. Image analysis was used to read and quantify the colorimetric signal and measure bacteria concentration, thus rendering this paper based microfluidic device quantitative. The colorimetric results show linearity over a wide concentration range (102CFU/mL to 108CFU/mL) and a limit of detection (LOD) of 103CFU/mL and 102CFU/mL for E. coli O157:H7 and S. Typhimurium, respectively. An insignificant change in colorimetric response for non-target bacteria indicates the aptasesnors are specific. The reported multiplexed colorimetric paper-based microfluidic devices is likely to perform well for on-site rapid screening of pathogenic bacteria in water and food products.