Overview

• LOCA-PRAM creators Han Lee and Brian Cunningham integrated deep learning with photonic resonator absorption microscopy to automate molecular biomarker analysis at the point of care.
• The team trained a physically grounded neural network by matching PRAM images of gold nanoparticle-tagged biomarkers against high-resolution scanning electron microscopy data.
• Benchmark tests showed LOCA-PRAM detects lower biomarker concentrations and reduces both false-positive and false-negative rates compared with standard image-analysis methods.
• Supported by the National Institutes of Health, the USDA AFRI Nanotechnology grant and the National Science Foundation, the research now moves toward clinical validation.
• The portable assay could empower patients and non-specialists to perform rapid, high-precision diagnostics for cancer, infections and treatment monitoring outside central labs.