Computational and machine learning approach for nanoparticles-enhanced bio-heat transport with coupled effects of interparticle spacing and particles radius.

Targeted delivery of drugs and hyperthermia in cardiovascular disease demand the accurate delivery of nanoparticles in complex arterial geometries. This paper introduces combined hybrid computational model that concomitantly examines the combined impact of nanoparticle radius and interparticle spacing on the thermal and mass transport characteristics of ternary bio-nanofluid flow under magnetohydrodynamic (MHD) effect. The ternary fluid is composed of blood fluid with suspended nanoparticles such as gold (Au), silver (Ag) silica SiO2. The mathematical model accounts for geometric properties of nanoparticles such as nanoparticles radius and interparticle spacing for their practical utility for several medical interventions. The numerical analysis is based on hybrid computational strategy, where the solutions are determined through the bvp4c numerical solver and then a novel supervised multi hidden layers Artificial neural network (ANN) is integrated. The proposed model has a high predictive capability with an exceptionally high accuracy with the lowest Mean squared error and ideal regression coefficient MSE=9.6327×10-11, Gradient=9.5681e-08, Mu=1e-09, and R2=1.0. Some of the main findings indicate that less spacing between particles (h=0.1) leads to continuous networks of thermal percolation, which enhance the thermal conductivity . to improve the efficiency of hyperthermia, whereas the larger nanoparticles radius .) offer a higher drug-loading capacity. .. Optimization of the magnetic parameter (.) also decreases flow velocity . and extends the nanoparticle residence time . which allows sustained drug delivery, results directly applicable to clinical-strength and . MRI-guided interventions. Radiation parameter . increases temperature of the ternary hybrid bio-nanofluids by . giving controllable thermal modulation to applications of hyperthermia. . The purpose of such an integrated computational-machine learning systems is to provide a better pathway, which could be crucial in several bio-medical interventions. The results can be used to offer practical advice to stent manufactures, interventional radiologist and pharmaceutical developers to create evidence-based cardiovascular therapy of the next generation.