Discerning Amyloid-β and Tau Pathologies with Learning-Based Quantum Sensing

Photon entanglement, a key feature of quantum correlations, provides a level of coherence absent in classical correlations, potentially offering new information when interacting with biological matter. One promising application is using entanglement decoherence to distinguish between healthy and diseased samples. However, achieving this requires efficient entangled photon sources capable of surviving through biological samples for reliable detection. In this work, we show the applicability of a polarization-entangled photon source as a label-free diagnostic tool for distinguishing between transgenic mouse models of amyloidosis and tauopathy and their respective control strains. We investigated cortical and hippocampal regions of these models, and our findings revealed greater preservation of entanglement in the transgenic samples compared to controls. To further enhance classification accuracy, we employed a supervised machine learning approach, achieving reliable distinctions between disease and control groups in unseen test samples. The quantum-based results were further validated through confocal imaging of the transgenic and control samples. These findings suggest that quantum sensing could serve as a label-free approach for distinguishing biological samples, with potential applications in the study of neurodegenerative disorders.