Interpretable AI-enabled electrochemical biosensors: Dendritic Fe?O?/graphene nanocomposites for next-generation clinical diagnostics

The synergetic combination of advanced nanomaterials and artificial intelligence (AI) is revolutionising the electrochemical biosensing field to achieve highly sensitive, selective and data-driven analytical platforms for clinical diagnostics. This keynote presentation highlights recent progress in the development of intelligent electrochemical sensors made from functional nanomaterials, with particular focus on dendritic Fe2O3/reduced graphene oxide (Fe2O3/rGO) nanocomposites for biomarker detection. A dendritic Fe2O3/rGO nanocomposite was prepared using a one-pot hydrothermal method and applied for the fabrication of a glassy carbon electrode for uric acid (UA) detection. The heterostructure was fully characterised by X-ray diffraction (XRD), Raman, field-emission scanning electron microscopy (FESEM), high-resolution transmission electron microscopy (HRTEM) and X-ray photoelectron spectroscopy (XPS) which confirmed the successful formation of the heterostructure with strong interfacial interactions. The synergistic effect of dendritic Fe2O3 and conductive graphene greatly improved the kinetics of electron transfer, the accessibility of the electrolyte, and the availability of electroactive sites, resulting in excellent electrocatalytic performance. The sensor showed a wide linear detection range of 1-70 μM, a low detection limit of 0.307 μM, a high sensitivity of 5.71 μA μM-1 cm-2, and excellent stability, reproducibility, and real-sample recovery (98.82%). Interpretable machine learning models were combined with differential pulse voltammetry (DPV) derived features to enhance the analytical reliability. Gaussian Process Regression and Ridge Regression showed high predictive accuracy (R2 = 0.95) and SHAP analysis identified peak current as the most influential parameter governing UA prediction. The convergence of nanotechnology, electrochemical sensing and explainable AI provides a powerful paradigm to develop robust, intelligent and clinically translatable biosensors for next generation healthcare applications. 

Keywords: Electrochemical biosensors; Fe2O3/reduced graphene oxide; Hydrothermal synthesis; Uric acid; Explainable artificial intelligence; Machine learning.

Dr. Manjunatha C is an Associate Professor of Chemistry at RV College of Engineering, Bengaluru, India, and Coordinator of the Centre of Excellence in Nanomaterials and Devices. His research focuses on functional nanomaterials for lithium-ion batteries, supercapacitors, hydrogen generation, photocatalysis, electrochemical sensors, and AI-enabled intelligent sensing. He has authored over 120 peer-reviewed publications, three book chapters, and holds six patents. Recognized among Stanford University's World's Top 2% Scientists (2023), he is a member of RSC (UK), ECS (USA), and ACS (USA) and actively collaborates with leading international research institutions in advanced materials and energy technologies.