Simultaneous Detection of Trace Protein Biomarkers from a Single Drop of Blood Using AI-Enhanced Smartphone-Based Digital Microscopy

The detection of early-stage diseases is often impeded by the low concentrations of protein biomarkers, necessitating sophisticated and costly technologies. In response, we have developed an advanced cyber-physical system that integrates blood plasma separation, biomarker detection, and analysis on a single microfluidic platform. This novel system enables the simultaneous detection of Neuron-Specific Enolase (NSE) and Carcinoembryonic Antigen (CEA) with high specificity and accuracy. Functionalized carbon dots (CDs) are employed for fluorescence-based quantification due to their superior biocompatibility, photostability, and resistance to photobleaching. The emission properties are optimized at 460 nm (blue) and 580 nm (yellow), yielding robust quantum efficiencies. The precise synthesis of CDs ensures reproducible fluorescence response with minimal background interference. A portable, smartphone-based fluorescence microscope equipped with 1000X magnification and UV excitation facilitates high-resolution image acquisition, serving as a low-cost alternative to conventional microscopy. Artificial intelligence algorithms are integrated for automated image analysis, enabling accurate quantification of biomarker concentrations. The system demonstrates impressive detection limits of 0.4 ng/mL and 0.9 ng/mL for CEA and NSE, respectively. The entire assay workflow, from sample introduction to result generation, is completed in 10 min. This integrated, portable diagnostic platform offers a transformative solution for point-of-care biomarker detection, particularly in resource-constrained settings, with the potential to democratize early disease screening and significantly reduce healthcare burdens globally.