Artificial intelligence based design of 3D-printed tablets for personalised medicine

Abstract

A multi-material 3D printing offers nearly endless possibilities for the spatial arrangement of individual materials within the object being printed. In the case of pharmaceutical tablets, the spatial arrangement of individual material domains containing the active pharmaceutical ingredients (APIs) and excipients uniquely defines the release profiles of the APIs. However, the inverse is not necessarily true – identical or very similar dissolution profiles can potentially be obtained from different tablet internal structures, implemented as a combination of domains containing excipients with different individual dissolution rates and different local API concentration. This work presents a computational method based on an Evolutionary Algorithm for the solution of the inverse problem, i.e. finding such tablet internal structure that results in a prescribed dissolution profile of each API contained in the tablet. After testing the algorithm on cases with a known solution, the methodology is applied to a problem of finding tablet structures that result in delayed release and step-wise release profiles, respectively. When combined with patient-specific requirements on drug release profiles, the algorithm can serve as a tool for an automated design of 3D-printed tablets in the framework of personalised medicine.