ABSTRACT
Spatiotemporal (ST) graph modeling has garnered increasing attention recently. Most existing methods rely on a predefined graph structure or construct a single learnable graph throughout training. However, it is challenging to use a predefined graph structure to capture dynamic ST changes effectively due to evolving node relationships over time. Furthermore, these methods typically utilize only the original data, neglecting external temporal factors. Therefore, we put forward a novel time-varying graph convolutional network model that integrates external factors for multifeature ST series prediction. Firstly, we construct a time-varying adjacency matrix using attention to capture dynamic spatial relationships among nodes. The graph structure adapts over time during training, validation, and testing phases. Then, we model temporal dependence by dilated causal convolution, leveraging gated activation unit and residual connection. Notably, the prediction accuracy is enhanced through the incorporation of embedding absolute time and the fusion of multifeature. This model has been applied to three real-world multifeature datasets, achieving state-of-the-art performance in all cases. Experiments show that the method has high accuracy and robustness when applied to multifeature and multivariate ST series problems.
ABSTRACT
BACKGROUND: Helicobacter pylori eradication with therapies employing a proton pump inhibitor (PPI) and antimicrobial agents is mainly achieved via bacterial susceptibility to antimicrobial agents and the magnitude of acid secretion inhibition. However, annual eradication rates have greatly declined in Mainland China, and therefore, tailored H. pylori eradication regimens that inhibit acid secretion and employ optimal antimicrobial agents determined based on gene clip testing may improve eradication rates. This study was aimed at evaluating the efficacy of tailored H. pylori eradication therapy guided by visual gene clip testing for antibiotic resistance and PPI metabolism genotypes. METHODS: This prospective study included 244 patients (141 men and 103 women aged 20-79 years) receiving initial treatment for H. pylori infection. Visual gene clip testing using gastric mucosal specimens was performed to detect antibiotic resistance to clarithromycin conferred by the A2142G and A2143G point mutations of the H. pylori 23S rRNA gene and to levofloxacin conferred by the Asn87 and Asp91 point mutations of the H. pylori gyrA gene. Patients received a 14-day bismuth quadruple therapy regimen guided by testing for antibiotic resistance and CYP2C19 polymorphisms, and primary H. pylori eradication was assessed at least 4 weeks after therapy. RESULTS: H. pylori strains were successfully isolated from the gastric mucosa tissues of 244 patients. Antibiotic resistant isolates were identified in 63 patients, with clarithromycin resistance observed in 50 patients, levofloxacin resistance in 7 patients, and dual resistance in 6 patients. The PPI metabolic genotype of CYP2C19 was detected in 242 of 244 cases, and 97 cases were categorized as extensive metabolizers, 141 as intermediate metabolizers, and 4 as poor metabolizers. Among the 242 patients who received tailored therapy, the H. pylori eradication rate was 90.9% (95% confidence interval 87.3%~94.6%) in the intention-to-treat analysis and 96.9% (95% confidence interval 94.7%~99.2%) in the per protocol analysis. CONCLUSIONS: Tailored therapy for H. pylori infection guided by determination of antibiotic resistance and CYP2C19 polymorphism using visual gene chip technology may provide high clinical effectiveness as initial H. pylori eradication therapy.