Machine-Learning-Based Random Lasing Diagnostics for Early Detection of Huntington Disease

Random Lasing (RL) is based on the interaction of an optical amplifying system and a highly scattering medium that produces the optical feedback necessary for laser emission. RL has demonstrated its potential as a diagnostic tool in biomedicine when biological tissues are used as the scattering source. Growing interest now focuses on the development of non-invasive techniques for the early detection of neurodegenerative diseases. In this study, we evaluated the early diagnostic potential of RL in a mouse model (R6/1 mice) of Huntington’s disease (HD), a neurodegenerative disorder caused by a CAG trinucleotide repeat expansion in the huntingtin gene. RL spectra were obtained from blood samples of 8-week-old R6/1 mice, prior to the onset of overt motor symptoms, and from their wild-type littermates. Multivariate statistical classification of the RL data yielded a moderate discrimination performance (~66% accuracy). However, after mathematical treatment of RL spectra and the application of machine-learning algorithms, the classification accuracy increased to ~90%. These findings support further investigation of RL signals from blood combined with machine-learning-based data as a potential non-invasive approach for the early detection of HD.