AI-Integrated Design and Optimization of Aceclofenac Loaded Transethosomal Gel

AbstractObjective:This study focused on the development and evaluation of aceclofenac-loaded transethosomes for improved topical delivery,along with the application of artificial intelligence, particularly the Random Forest Regression (RFR) model, to predict andoptimize formulation performance.Methods:Transethosomes containing aceclofenac were prepared using the thin-film hydration method with phosphatidylcholine,ethanol, and surfactants (Tween 80 and Span 80). The RFR algorithm was employed to establish a relationship betweenformulation variables and key responses such as entrapment efficiency (EE%) and cumulative drug release (CDR%). Drug–excipient compatibility was assessed using FT-IR analysis. The optimized vesicles were characterized for entrapmentefficiency, particle size, zeta potential and morphology. The formulation was incorporated into a Carbopol 934 gel andevaluated for physicochemical properties and In vitro drug release. Stabilitystudies were conducted to examine formulationintegrity over time.Results:The optimized transethosomal gel showed nanoscale vesicles with good stability characteristics. The RFR model effectivelypredicted formulation behaviour and aided in selecting optimal parameters. FT-IR results indicated no significant interactionbetween the drug and excipients. The optimized formulation demonstrated an entrapment efficiency of 78.38% along with acontrolled and sustained drug release pattern. Stability testing confirmed that the formulation maintained its propertiesthroughout the study duration.Conclusion: The study successfully demonstrated that Random Forest Regression can be effectively utilized to optimize aceclofenac-loaded transethosomal formulations. The developed system showed desirable stability and sustained drug release, indicating its suitability for enhanced topical drug delivery.