Understanding the complexity of sepsis mortality prediction via rule discovery and analysis: a pilot study

Background Sepsis, defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, has become one of the major causes of death in Intensive Care Units (ICUs). The heterogeneity and complexity of this syndrome lead to the absence of golden standards for its diagnosis, treatment, and prognosis. The early prediction of in-hospital mortality for sepsis patients is not only meaningful to medical decision making, but more importantly, relates to the well-being of patients. Methods In this paper, a rule discovery and analysis (rule-based) method is used to predict the in-hospital death events of 2021 ICU patients diagnosed with sepsis using the MIMIC-III database. The method mainly includes two phases: rule discovery phase and rule analysis phase. In the rule discovery phase, the RuleFit method is employed to mine multiple hidden rules which are capable to predict individual in-hospital death events. In the rule analysis phase, survival analysis and decomposition analysis are carried out to test and justify the risk prediction ability of these rules. Then by leveraging a subset of these rules, we establish a prediction model that is both more accurate at the in-hospital death prediction task and more interpretable than most comparable methods. Results In our experiment, RuleFit generates 77 risk prediction rules, and the average area under the curve (AUC) of the prediction model based on 62 of these rules reaches 0.781 ($$\pm 0.018$$ ± 0.018 ) which is comparable to or even better than the AUC of existing methods (i.e., commonly used medical scoring system and benchmark machine learning models). External validation of the prediction power of these 62 rules on another 1468 sepsis patients not included in MIMIC-III in ICU provides further supporting evidence for the superiority of the rule-based method. In addition, we discuss and explain in detail the rules with better risk prediction ability. Glasgow Coma Scale (GCS), serum potassium, and serum bilirubin are found to be the most important risk factors for predicting patient death. Conclusion Our study demonstrates that, with the rule-based method, we could not only make accurate prediction on in-hospital death events of sepsis patients, but also reveal the complex relationship between sepsis-related risk factors through the rules themselves, so as to improve our understanding of the complexity of sepsis as well as its population.

Oral Administration of East Asian Herbal Medicine for Peripheral Neuropathy: A Systematic Review and Meta-Analysis with Association Rule Analysis to Identify Core Herb Combinations

This review aimed to comprehensively assess the efficacy and safety of oral East Asian herbal medicine (EAHM) for overall peripheral neuropathy (PN). In addition, an Apriori algorithm-based association rule analysis was performed to identify the core herb combination, thereby further generating useful hypotheses for subsequent drug discovery. A total of 10 databases were searched electronically from inception to July 2021. Randomized clinical trials (RCTs) comparing EAHM with conventional analgesic medication or usual care for managing PN were included. The RCT quality was appraised using RoB 2.0, and the random effects model was used to calculate the effect sizes of the included RCTs. The overall quality of evidence was evaluated according to the Grading of Recommendations Assessment, Development, and Evaluation. By analyzing the constituent herb data, the potential association rules of core herb combinations were explored. A total of 67 RCTs involving 5753 patients were included in this systematic review. In a meta-analysis, EAHM monotherapy and combined EAHM and western medicine therapy demonstrated substantially improved sensory nerve conduction velocity, motor nerve conduction velocity, and response rate. Moreover, EAHM significantly improved the incidence rate, pain intensity, Toronto clinical scoring system, and Michigan diabetic neuropathy score. The evidence grade was moderate to low due to the substantial heterogeneity among the studies. Nine association rules were identified by performing the association rule analysis on the extraction data of 156 EAHM herbs. Therefore, the constituents of the herb combinations with consistent association rules were Astragali Radix, Cinnamomi Ramulus, and Spatholobi Calulis. This meta-analysis supports the hypothesis that EAHM monotherapy and combined therapy may be beneficial for PN patients, and follow-up research should be conducted to confirm the precise action target of the core herb.

An Association Rule Analysis of the Acupressure Effect on Sleep Quality

Background. Sleep is recognized as an all-important physiological process, which also contributes to maintaining several bodily functions and systems. According to the Pittsburgh Sleep Quality Index (PSQI), also known as the most widely used tool in the field of subjective assessment of self-perceived sleep quality, a combination of acupoints could be more effective than single acupoint treatment in improving sleep quality. Methods. The present study was based on the extracted eligible studies rooted in a previous meta-analysis that worked on the basis of association rule mining and examined the potential kernel acupoint combinations for improving sleep quality. Results. Depending on the Apriori algorithm, we summarized 26 acupoints as binary data from the 32 eligible studies based on a previous meta-analysis and analyzed them. The top 10 most frequently selected acupoints were HT7, SP6, PC6, KI1, GV20, EM5, EX-HN3, EX-HN16, KI3, and MA-TF1. Furthermore, as deduced from 21 association rules, the primary relevant rules in the combination of acupoints are (EX-HN3, EX-HN16)=>(GV20) and (HT7, KI1)=>(PC6). Conclusions. In order to use acupuncture to improve sleep quality, integrating (EX-HN3, EX-HN16, GV20) with (HT7, KI1, PC6) acupoints could be deemed as the kernel acupoint combination.

Analisa Pola Penjualan Produk Sepeda Motor Yamaha Menggunakan Metode Algoritma Apriori

The development of automotive industry in Indonesia can be classifiedas very rapid and annually increasing, causing highly competitive circumstances because many companies provide various types of motorcycle brands with quality and competitive prices. The company must create a marketing strategy pattern that can increase the level of sales efficiency of Yamaha motorcycle products. To overcome this problem, a strategy that can help increasing sales of motorcycle products is needed, in which by utilizing sales data owned by the company. Data mining can be used to process company sales data by looking for association rules with apriori algorithm on motorcycle product variables. From the results of the association rule analysis on sales data, with a minimum support of 30% and a minimum confidence of 75% can produce 3 rules with 3 products that are most in demand by consumers, namely the NEW MIOM3 CW, NEWAEROX155VVA and N-MAX, by knowing the most selling products, the company can add the most selling product supply and develop a marketing strategy to market the products with other products by examining the comparative advantage of the most sold products over the other products.