Smart design of intratumoral thermosensitive β-lapachone hydrogels by Artificial Neural Networks

This study presents Artificial Neural Networks (ANN) as a tool for designing injectable intratumoral formulations of the anticancer drug β-lapachone. This methodology permits insight into the interactions between variables and determines the design space of the formulation without the restrictions of an experimental design. An ANN model for two critical parameters of the formulations; the amount of solubilized drug and gel temperature was developed and validated. The model allowed an understanding of interactions between ingredients in the formulation and the fundamental phenomena as the formation of polypseudorotaxanes to be detected and quantified.