Additive Manufacturing Wearable Magnetic Sensors: Revolutionizing Cardiac Health Monitoring with Machine Learning for Arrhythmia Classification - PROJECT SUMMARY Magnetocardiography (MCG) measures the weak magnetic fields (1 - 100 picotesla, 0.1 - 100 Hz) generated by the heart’s electrical activities using superconducting quantum interference devices (SQUIDs). It complements electrocardiography (ECG) by providing greater detail, as MCGs are more sensitive to currents tangential to the chest surface and can detect vortex currents. Due to its high independence from inhomogeneities in electrical resistivity inside the torso and on the skin, MCG offers a practical alternative for monitoring various cardiac conditions. However, SQUIDs require cryogenic cooling, specialized magnetic shielding rooms, and regular maintenance, making them expensive and logistically challenging to operate. Additionally, their bulkiness and complexity limit portability and require specialized training, further restricting widespread use. This SuRE project aims to advance MCG technology by combining additive-manufactured organic granular magnetoresistive (OgMR) sensors with machine learning for superior cardiac diagnostics. Unlike traditional magnetoresistive (MR) sensors such as giant magnetoresistive (GMR), magnetic tunnel junction (MTJ), and anisotropic magnetoresistive (AMR) sensors, OgMR sensors incorporate magnetic granules (typically nanoparticles) into organic semiconductors, eliminating the need for thin film deposition and micro/nano-fabrication facilities, thus significantly reducing fabrication costs. Additionally, OgMR sensors can be directly printed on flexible substrates with consistent quality in high volumes. Our preliminary results indicate that OgMR sensors offer 50 times higher MR compared to GMR sensors and comparable performance to MTJs, making them a promising candidate for inexpensive, flexible, high-sensitivity magnetic sensors for cardiac signal recording. Given the advantages of OgMR sensors and the diverse expertise of the PI, Co-I, and consultant, we aim to achieve the following specific goals: (1) additive manufacturing and characterizing flexible OgMR sensors; (2) flexible OgMR sensor packaging and external circuit design; (3) flexible OgMR sensors for cardiac rhythm recording and machine learning for cardiac arrhythmias classification. We will develop a palm-sized circuit board with a 5 mm × 8 mm flexible OgMR sensor for ex-vivo and in-vivo mouse cardiac rhythm recording. By the end of this project, we expect to develop a portable OgMR sensing platform for MCG recording and a machine- learning algorithm for the high-accuracy classification of normal and arrhythmic heart rhythms. In the long run, our work aims to revolutionize the MCG technique and reduce healthcare costs.
