This paper utilizes response surface methodology (RSM) design-based analyses to optimize starch-based wound dressings that were characterized based on weight loss (WL%), swelling index (SI%), and mechanical strength (TS). The wound-dressing materials were prepared by employing a modified casting method, using various concentrations of starch (0.5-2 w/w%), polyvinyl alcohol (PVA) (0.5-2 weight%), citric acid (1.5-4 weight%), and glycerol (1.5-4 weight%) to yield wound-dressing films with appropriate combinations of in vitro degradation, swelling index, and tensile strength. As a result of the response surface method-based analysis, the swelling index, in vitro degradation, and tensile strength were linearly related to variations in the formulation of organic components. Based on our experimental investigations, the optimized film (formulation: 1 weight% PVA, 2 weight% starch, 1.5 weight% citric acid, and 1.5 weight% glycerol) exhibited an outstanding swelling index (343.52%), suitable in vitro degradation (53.22%), and excellent tensile strength (8.82 MPa). The response surface plots for the dependent variables, swelling index (SI%), weight loss (WL%), and mechanical strength (TS), showed that in all dual relations, the PVA-starch combination significantly affected all dependent variables; however, the PVA-citric acid interaction showed the most excellent effect on the swelling index. As a result, every component of the resulting film had a lesser amount of all ingredients to achieve better properties at a lower material cost. Starch-based/PVA films have been identified in this paper as optimal and more affordable wound-dressing films.
Keywords: PVA; biodegradable; machine learning; optimization; starch; wound dressings.