A RCA-based signal amplification and machine learning-assisted sandwich-type dual-mode split aptamers sensor for the...

Sensitive detection of diarrhetic shellfish toxins (DSTs) is vital for ensuring marine food safety. In this study, a novel sandwich-type dual-mode split aptamers sensor (DMSAS) was constructed for the ultrasensitive simultaneous detection of DSTs by integrating rolling circle amplification (RCA)-based signal enhancement, fluorescent/colorimetric dual-mode output, and machine learning (ML)-based intelligent data analysis. Under the optimal conditions, DMSAS displayed limits of detection of 4.510 fg/mL and 4.859 fg/mL for fluorescent and colorimetric detection modes, which were improved to 3.383 fg/mL after ML-based dual-signal fusion. An intelligent recognition model for DSTs concentrations that was established using ML could achieve 100% classification accuracy for DSTs in simulated natural shellfish samples. Spiked recovery experiments yielded recovery rates ranging from 88.9% to 103.1%, with relative standard deviations below 12%, indicating the satisfactory accuracy, repeatability, and matrix adaptability of DMSAS. This work provides a new strategy for monitoring marine food safety.