Feed

Liquid biopsy coupled with optical spectroscopy holds immense potential for disease screening. Specifically, laser-induced breakdown spectroscopy (LIBS) and fourier transform infrared spectroscopy (FTIR) provide rapid, complementary insight into elemental and molecular profiles. During biofluid drying, the "coffee-ring effect" (CRE) produces an uneven solute deposit, causing large fluctuations in LIBS signals. FTIR molecular features are generally less affected by this heterogeneity. We therefore designed a superhydrophobic SiO₂NPs@Si-microgroove and combined it with a multimodal CNN for renal cell carcinoma (RCC) detection. The substrate confined the urine droplet during drying, improved deposit uniformity, and enhanced LIBS measurement repeatability, reducing the relative standard deviation (RSD) of representative emission signals to approximately 10%. By combining the more reproducible LIBS measurements with complementary FTIR data, the CNN model achieved an Accuracy (ACC) of 99.22% for RCC detection. This work offers a novel approach for the standardized, intelligent, and precise analysis of future biofluids.