Loïc Van Dieren Vlad Tereshenko Haïzam Oubari Yanis Berkane Jonathan Cornacchini Filip Thiessen Curtis L. Cetrulo Korkut Uygun Alexandre Lellouch ^*

• Harvard University, Cambridge, United States

Introduction: Magne'c nanopar'cles (MNPs), par'cularly iron oxide nanopar'cles (IONPs), are renowned for their superparamagne'c behavior, allowing precise control under external magne'c fields. This characteris'c makes them ideal for biomedical applica'ons, including diagnos'cs and drug delivery. Superparamagne'c IONPs, which exhibit magne'za'on only in the presence of an external field, can be functionalized with ligands for targeted affinity diagnostics. This study presents a computational model to explore the induced voltage in a search coil when MNPs pass through a simulated blood vessel, aiming to improve non-invasive diagnos'c methods for disease detec'on and monitoring. Methods: A finite element model was constructed using COMSOL Multiphysics to simulate the behavior of IONPs within a dynamic blood vessel environment. Governing equa'ons such as Ampère's Law and Faraday's Law of Induc'on were incorporated to simulate the induced voltage in a copper coil as MNPs of various sizes flowed through the vessel. Rheological parameters, including blood viscosity and flow rates, were factored into the model using a non-Newtonian fluid approach. Results:The amount of MNPs required for detec'on varies significantly based on the sensi'vity of the detec'on equipment and the size of the nanopar'cles themselves. For highly sensi've devices like a SQUID voltmeter, with a coil sensi'vity around 10 -12 V, very low MNP concentra'ons -approximately 10 -4 µg/ml -are sufficient for detec'on, staying well within the safe range. As coil sensi'vity decreases, such as with standard voltmeters at 10 -8 V or 10 -6 V, the MNP concentra'on required for detec'on rises, approaching or even exceeding levels considered poten'ally toxic. Addi'onally, the physical size of MNPs plays a role; larger nanopar'cles (e.g., 50 nm radius) require fewer total par'cles for detec'on at the same sensi'vity, compared to smaller par'cles like those with a 2.5 nm radius. For instance, at a coil sensi'vity of 10 -10 V, a 2.5 nm par'cle requires around 10 12 par'cles, whereas a 50 nm par'cle only needs 10 8 . This highlights the importance of op'mizing both detec'on sensi'vity and par'cle size to balance effec've detec'on with safety. Conclusion: This computa'onal model demonstrates the feasibility of using superparamagne'c nanopar'cles in real-'me,non-invasive diagnos'c systems.