Related Experiment Video
Updated: Sep 13, 2026

Detection and Isolation of Cancer in Prostate Biopsies Using Stimulated Raman Histology and Artificial Intelligence
Published on: June 10, 2025
Dual-mode immunosensor combining quantitative surface-enhanced Raman scattering and visual upconversion luminescence
Panqing Xu1, Qing Xie1, Yang Yan1
1Department of Radiology, The Affiliated Yangming Hospital of Ningbo University, Ningbo, 315400, China.
Abstract:
Timely detection and precise clinical management are essential to improving outcomes in prostate cancer, which remains a leading global health concern. Herein, we developed an innovative optical sensor combining surface-enhanced Raman scattering (SERS)-based quantitative detection and upconversion luminescence (UCL)-mediated visual recognition for prostate-specific antigen (PSA). This negatively correlated dual-functionality specially depended on mulberry-like NaGdF4:Yb3 +,Er3+ nanoparticles with highly textured surfaces, which were successfully constructed as the capture substrate. Crucially, the intrinsic 4f-4f electronic transitions of the doped lanthanide ions not only generated stable UCL emissions but also resonated with the vibrational bands of the Raman reporter (4-mercaptopyridine). This energy level alignment activated a robust photoinduced charge transfer (PICT) channel, driving exceptional chemical enhancement. This synergistic physical enrichment and chemical enhancement of the immunosubstrate yielded an impressive initial SERS enhancement factor of 1.01 × 106. To construct the sandwich architecture, gold nanourchins (GNUs) were employed as plasmonic immunoprobes, whose proximity to the immunosubstrate triggered a unique inversely correlated optical response. The SERS signal was dramatically amplified via intense plasmonic coupling, accompanied by the proportional quenching of the UCL signal via non-radiative energy transfer. Consequently, this synergistic paradigm demonstrated unprecedented analytical sensitivity, achieving an ultralow limit of detection as 1.08 × 10-9 mg/mL alongside an intuitive naked-eye visualization. Finally validated by complex biological samples, this cross-validating sensing platform paves a highly reliable avenue for the rapid and precise screening of cancer biomarkers.

