Z.-T. Lin, C. Tang, Y. Kong, Y. Peng, N. Jiang, S. Zhou, Y. Zhang, B. Jana, S. Korupolu, S. Gao, Y. Li, Z. Wang, C. Liang, Q. Zhang, Y. Liang, R. Saxena, W. Shou, W. Matusik, C. L. Evans, T. Wu, M. X. Wu
Many diseases, particularly chronic diseases, involve complex biological processes that cannot be adequately assessed by a single biomarker. However, identifying a specific biomarker panel and implementing its measurement at point-of-care (PoC) or home testing remain highly challenging. Here, we developed a micro-ChromaDot array (McDa) platform for multiplexed biomarker measurement in a one-biomarker-one-dot fashion. The system engraved colorimetric signals from a 3D microneedle array into a 2D gel via a chem-spatial signal-amplification cascade, enabling precise smartphone-based signal acquisition and analysis. Using clinical sera from systemic lupus erythematosus (SLE) patients as a disease model, we quantified five biomarkers and integrated the data with deep-learning-assisted modeling. A support vector machine (SVM) model with 5-fold cross-validation identified an optimized three-biomarker panel that discriminated SLE patients from healthy controls with 97% specificity and 93% sensitivity, substantially outperforming the current clinical standard antinuclear antibody (ANA) test. To translate the biomarker-panel analysis into decentralized testing, we further engineered a mini-diagnostic device (miniDia) and associated App capable of automated McDa assays, smartphone-based analysis, and patient-friendly result readout without requiring a laboratory or trained personnel. This portable, potentially home-use platform achieved sensitivity superior to conventional enzyme-linked immunosorbent assays (ELISA). This adaptable and cost-effective platform holds strong potential for home-based diagnosis, prognosis, and monitoring of various chronic diseases.