Evaluating the effect of griseofulvin on the disintegration properties of tablets.

Understanding how active pharmaceutical ingredients (APIs) influence tablet disintegration remains essential for optimising oral drug product performance. In this work, we present an integrated, data-driven framework that combines deep-learning-based image segmentation, automated regime detection and mechanistic modelling to characterise liquid absorption and swelling behaviours from sessile drop experiments. Using this workflow, we investigated the effect of incorporating griseofulvin into directly compressed tablets containing a variety of excipient combinations. Across all formulations, the presence of griseofulvin significantly reduced wettability and maximum swelling size, consistent with replacing hygroscopic excipients with a poorly soluble, hydrophobic drug. Bayesian hierarchical modelling confirmed credible decreases in swelling capacity for all formulations and formulation-dependent changes in initial swelling rate. Notably, MCC/lactose tablets exhibited faster initial swelling and shorter disintegration times despite reduced porosity, suggesting API-induced microstructural changes that enhance access of liquid to the swellable excipients. Overall, the proposed framework provides a rapid (<6 min), mechanistic and minimally subjective method for probing formulation-dependent disintegration behaviour. For griseofulvin, the method was sensitive to changes in absorption and swelling behaviour, highlighting its potential as a high-throughput formulation screening approach. Further studies involving a wider range of APIs are needed to determine whether these observations are generalisable and relevant to disintegration and dissolution performance.