3D Printing of chewable oral tablets using drug nanosuspension inks: an experimental and machine learning study

3D printing enables precise control over tablet design and drug release, but challenges remain in optimising ink formulation, ensuring printability, and predicting final tablet properties. This study addresses the need for data-driven strategies in fabricating chewable tablets and tests the hypothesis that integrating rheology with machine learning (ML) enables predictive control over print quality and dosage form performance. We developed drug nanosuspension inks with varying water content (85–20 wt.%) and identified 40% as optimal, balancing shear-thinning behaviour, yield stress, and shear recovery for consistent extrusion. Analytical models predicted strut diameter (D) based on printing parameters—pressure (P), speed (v), and nozzle height (h)—but showed reduced accuracy under nonideal conditions. A Gradient Boosting Regressor (GBR) model improved predictions (R² = 0.94, RMSE = 52 µm) and enabled inverse prediction of printing settings for target dimensions. Tablets were characterised for drug stability, content uniformity, mechanical properties, and dissolution. XRD and DSC confirmed preserved GF crystallinity, and Raman and UV-VIS analysis demonstrated uniform drug distribution. Tablets had a stiffness of ∼74 kPa, similar to commercial gummies, and dissolution rates were tunable via infill design. This work establishes a predictive, rheology-informed, and ML-enhanced framework for the scalable and customisable 3D printing of chewable oral tablets.