Lyme disease (LD) is a tick-borne illness and can lead to severe health complications if left undiagnosed/untreated. This study introduces a novel approach to point-of-care testing for LD, utilizing deep-learning-enabled peptide-based serodiagnosis. 82 unique patient serum samples were used for validation, obtained from the Bay Area Lyme Disease Biobank and the CDC Lyme Serum Repository. Blinded test results demonstrated outstanding diagnostic performance, with 84.0% sensitivity and 96.3% specificity. By integrating deep learning techniques with a peptide-based sensing panel, this serodiagnosis offers a rapid (< 15 minutes) and cost-effective (< $0.5/test) platform for LD diagnostics, while minimizing cross-reactivity.
|