Assessing the Adequacy of Hemodialysis Patients via the Graph-Based Takagi-Sugeno-Kang Fuzzy System

The ability of phase angle and body composition to predict risk of death in maintenance hemodialysis patients

OBJECTIVE

The objective of this study was to investigate the ability of phase angle and body composition to identify risk factors for mortality among patients receiving maintenance hemodialysis (MHD) treatment.

METHODS

In this retrospective study, we examined the causes of death in 43 MHD patients who were treated at our hemodialysis center between January 2016 and December 2021 and compared the patients to 71 patients who survived during the same period. Body composition was measured using direct segmental multi-frequency bioelectrical impedance to obtain phase angle, fat-free mass (FFM), extracellular water/total body water (ECW/TBW), and waist circumference (WC). Laboratory data were also collected. Phase angle cut-off value-associated variables were identified using ROC analysis. The ability of body composition variables to identify risk factors for death in MHD patients was evaluated.

RESULTS

We found that cardiovascular disease was the most common cause of death among MHD patients. ROC curve analysis revealed that the optimal cut-off value for phase angle as a predictor of death risk in MHD patients was 4.50°. Additionally, lower phase angle, increased age, longer dialysis vintage, lower KT/V, and hypoproteinemia were identified as significant risk factors for death in MHD patients.

CONCLUSION

In conclusion, our findings suggest that cardiovascular disease is the leading cause of death among MHD patients and that lower phase angle, increased age, longer dialysis duration, and hypoproteinemia can be used to predict the risk of mortality in this patient population. The underlying mechanism by which lower phase angle can be used to predict the prognosis of MHD patients warrants further investigation.

A hybrid Genetic–Grey Wolf Optimization algorithm for optimizing Takagi–Sugeno–Kang fuzzy systems

Nature-inspired optimization techniques have been applied in various fields of study to solve optimization problems. Since designing a Fuzzy System (FS) can be considered one of the most complex optimization problems, many meta-heuristic optimizations have been developed to design FS structures. This paper aims to design a Takagi–Sugeno–Kang fuzzy Systems (TSK-FS) structure by generating the required fuzzy rules and selecting the most influential parameters for these rules. In this context, a new hybrid nature-inspired algorithm is proposed, namely Genetic–Grey Wolf Optimization (GGWO) algorithm, to optimize TSK-FSs. In GGWO, a hybridization of the genetic algorithm (GA) and the grey wolf optimizer (GWO) is applied to overcome the premature convergence and poor solution exploitation of the standard GWO. Using genetic crossover and mutation operators accelerates the exploration process and efficiently reaches the best solution (rule generation) within a reasonable time. The proposed GGWO is tested on several benchmark functions compared with other nature-inspired optimization algorithms. The result of simulations applied to the fuzzy control of nonlinear plants shows the superiority of GGWO in designing TSK-FSs with high accuracy compared with different optimization algorithms in terms of Root Mean Squared Error (RMSE) and computational time.