Multivariate predictive modeling of photothermal response of indocyanine green loaded graphene oxide under continuous and pulsed laser irradiation

The integration of photothermal therapy (PTT) into a single nanoplatform is a promising approach for precise cancer treatment; however, achieving predictable and optimized thermal doses remains a key challenge. While previous studies have explored graphene oxide (GO) and indocyanine green (ICG) composites, this work introduces a data-driven strategy to model and predict the photothermal response. We developed an ICG-GO nanocomposite and systematically investigated its performance under 808 nm near-infrared pulsed and continuous-wave laser irradiation. ICG functionalization enhanced the photothermal response of GO by up to 2.5-fold, achieving a rapid temperature increase to 45 °C within 2 min compared to 5 min for GO alone at 0.3 mg/mL and 1.5 W. Pulsed irradiation (laser-on: 200–1000 ms; laser-off: 50–150 ms) was identified as a crucial parameter for optimizing thermal management, reducing unwanted heat diffusion while maintaining therapeutic efficacy. Biocompatibility assessments confirmed that the integration of ICG attenuated the inherent cytotoxicity of GO, and the ICG-GO platform enabled efficient cancer cell ablation (ICG-GO at a concentration of 0.3 mg/mL, when exposed to continuous-wave laser irradiation at 1.5 W for 5 min, resulted in 98% cell death (2% cell viability)). The core innovation of this study lies in the application of multivariate statistical analysis, including ANOVA and regression modeling, which generated a predictive equation for the temperature output based on key laser parameters. This model provides a quantitative framework for optimizing PTT protocols. Thus, this study establishes the ICG-GO nanocomposite as an experimentally characterized photothermal platform and presents a predictive statistical framework that may facilitate future optimization of photothermal treatment parameters in preclinical research.