AI-driven entropy analysis of Sutterby tetra-hybrid nanofluid energy conversion for electromagnetic peristaltic flow

Predicting nano-blood transport in electromagnetically actuated peristaltic microchannels requires solving coupled nonlinear equations for momentum, energy, mass, and microorganism conservation a computationally expensive task when Hall and ion-slip effects, thermal radiation, viscous dissipation, entropy generation, and Sutterby non-Newtonian rheology are all included, limiting real-time clinical use. This study addresses that bottleneck by developing a Levenberg-Marquardt optimized artificial neural network (ANN-LM) trained on 2,400 high-fidelity datasets generated from the governing Multiphysics equations, to predict velocity, temperature, concentration, and microorganism distributions for a Sutterby tetra-hybrid nanofluid. The ANN-LM achieves mean absolute errors below 0.3% and correlation coefficients near unity across all fields, while reducing computational cost by more than two orders of magnitude relative to direct numerical solution. By coupling deep learning with electromagnetic-rheological-bioconvective modeling for the first time in this configuration, the framework enables millisecond-scale parametric exploration and inverse design previously infeasible with conventional solvers offering a practical route toward real-time, patient-specific control in magnetically guided drug delivery, cardiovascular interventions, and microfluidic diagnostic platforms.