Noise-reduced optical detection of mirna-155 by porous silicon through morlet wavelet phase analysis

Sensitive and selective detection of miRNA-155 was achieved with the nanostructured pSi-based optical biosensor in PBS under controlled conditions, free from matrix effects by analyzing Fabry–Perot fringes obtained from reflectance measurements using MWPA signal processing method. MWPA clearly outperformed RIFTS because frequency and phase components were effectively separated and low-frequency noise was suppressed, resulting in consistent and precise ΔEOT measurements across the entire data set. In contrast, only approximately 28% of the data were reliably processed by RIFTS, while the remaining measurements were inconsistent and affected by noise. The present findings indicate that the reliable detection of low-concentration miRNA in pSi thin-film biosensors is achieved through the combination of nanostructure optimization and advanced signal processing. The sensor LOD was 0.13 pM with MWPA, while RIFTS yielded ∼0.72 pM with low reliability. Reproducibility was confirmed across chips from three independently fabricated, identical wafers, with triplicate measurements at each concentration yielding consistent sensor responses. Reusability is retained after denaturation, with minor performance loss. Selectivity was demonstrated using miRNA-148, while testing additional sequences exceeds the scope of this work and provides a basis for future investigations. The work demonstrates a proof-of-concept pSi thin-film miRNA sensor in PBS with an LOD ∼ six times lower than RIFTS, providing a basis for future testing in complex physiological fluids via microfluidic integration and smartphone-based readout.