Integrating machine-learning and nanotechnology: imatinib-loaded nanomicelles for targeted therapy in MCF-7 breast cancer

Using machine-learning (ML) algorithms, we developed and characterized imatinib-loaded nanomicelles to enhance tumor-targeted delivery and reduce systemic toxicity. Nanomicelles were synthesized via an F127-mediated oil-in-water self-assembly process, achieving 65 % encapsulation efficiency and a uniform particle size of 257 nm. In-vitro studies showed pronounced pH-responsive release, with accelerated drug liberation at acidic pH (5.4) mimicking the tumor microenvironment and sustained release at physiological pH (7.4). To analyse release kinetics, we applied a supervised ML framework across eight regression algorithms. XGBoost delivered the best accuracy (R 2 = 0.998, MAE = 0.87) with strong cross-validation (CV, R 2 = 0.994). Modeling revealed 38–42 % faster release at pH 5.4, with time–pH interaction explaining 72 % of kinetic variability. These in-vitro results suggest that nanomicelles enable pH-triggered release and that ML can predict release kinetics, warranting further validation in biological systems. • Imatinib-loaded nanomicelles show high encapsulation efficiency, uniform size & pH-responsive release for targeted therapy. • ML algorithms, including XGBoost, accurately predict complex pH-dependent drug release kinetics. • Nanotechnology and ML integration offers mechanistic insights for rational tumor-targeted drug design.