application of artificial neural network and response surface methodology approach for modeling in-vitro removal of cu(ii) ion from human blood plasma using a medicinal biomass

Industrial activities have recently propelled the excessive introduction of toxic metals into the environment, leading to diverse health challenges. therefore, it is apt to detoxify the blood transport system (blood plasma). the present study employed a batch technique to conduct in-vitro cu(ii) removal from human blood plasma onto opuntia fragalis leaves (ofl). the adsorption process was optimized utilizing response surface methodology (rsm) and artificial neural network (ann) models. the 3-d response surface graph was utilized to identify the interaction influence of key factors on the percentage removal of cu(ii) ion. the adsorbent dose (1.5 mg), initial concentration (30 mg/l), ph (6.03), and contact duration (65 mins) were proposed as the best parameters using the ann prediction profiler at 98.45 % removal of cu(ii) ion from blood plasma which is similar to rsm values. the quadratic model demonstrated an excellent fit for the experimental data with a coefficient of correlation (r2) of 0.8627 and an f-value of 7.73. the ann model derived from the same design demonstrated adequate predictive performance of cu(ii) ion removal, which was reasonably predicted with an excellent correlation between the predicted and experimental values (r2 = 0.9160). the developed models were evaluated utilizing the root mean square error (rmse), and coefficient of correlation (r2) to identify the best ann topology. furthermore, the algorithm with the lowest rmse (0.241) and highest r2 was observed to be trained best using the levenberg-marquardt algorithm. additionally, ann results were deemed more reliable than rsm results and ofl has demonstrated excellent cu(ii) ion removal from the human blood plasma.