Denoising Algorithm for High-Resolution and Large-Range Phase-Sensitive SPR Imaging Based on PFA

Abstract

Phase-sensitive surface plasmon resonance (SPR) detection is widely employed in molecular dynamics studies and SPR imaging owing to its real-time capability, high sensitivity, and compatibility with imaging systems. A key research objective is to achieve higher measurement resolution of refractive index under optimal dynamic range conditions. We present an enhanced SPR phase imaging system combining a quad-polarization filter array for phase differential detection with a novel polarization pair, block matching, and 4D filtering (PPBM4D) algorithm to extend the dynamic range and enhance resolution. By extending the BM3D framework, PPBM4D leverages inter-polarization correlations to generate virtual measurements for each channel in the quad-polarization filter, enabling more effective noise suppression through collaborative filtering. The algorithm demonstrates 57% instrumental noise reduction and achieves 1.51 × 10^-6 RIU resolution (1.333-1.393 RIU range). The system's algorithm performance is validated through stepwise NaCl solution switching experiments (0.0025-0.08%) and protein interaction assays (0.15625-20 μg/mL). This advancement establishes a robust framework for high-resolution SPR applications across a broad dynamic range, particularly benefiting live-cell imaging and high-throughput screening.

Keywords: BM3D; SPR imaging; denoising algorithm; phase-sensitive SPR; polarization filter array.

Figure 1

(a). Schematic illustration of the optical configuration based on PFA. (b). Basic principle of PFA. (c). Photograph of the experimental setup. (d). Photograph of detailed microfluidics chip and prism structures.

Figure 2

The workflow of the phase image extraction principle.

Figure 3

The flowchart of the PPBM4D algorithm.

Figure 4

The flowchart of experimental protocols. (a). Protocol of glycerol-deionized water solutions switching. (b). Protocol of NaCl-deionized water solution switching. (c). Protocol of protein binding kinetics measurement.

Figure 5

Linear range and sensitivity calibration of the measurement system. (a). Theoretical curve of phase and RIRF responses versus refractive index. Blue curve (left axis) displays the RIRF theoretical response with a linear range (R² @99%) of 1.340–1.376, while the red curve (right axis) shows the phase response with an extended linear range (R² @99%) of 1.333–1.393 RIU. The experimental data (red circles) exhibit excellent agreement with the theoretical phase curve. (b). Time-resolved phase response during glycerol concentration gradient testing, showing characteristic step-like transitions. The refractive index is marked besides the curve. (c). Linear regression of averaged phase values versus glycerol concentrations yields a system sensitivity of 52.504 rad/RIU (R² = 0.9965).

Figure 6

Spatial denoising effect of phase images. (a) Ground truth phase image. (b) Typical raw phase image. Detailed regions are shown below with pseudo-color difference maps relative to ground truth. (c) Denoised phase image. Detailed regions are shown below with pseudo-color difference maps relative to ground truth. (d) Pixel-wise phase error histogram of raw single-frame measurement. (e) Pixel-wise phase error histogram of denoised single-frame measurement.

Figure 7

Resolution enhancement validation through dilute solution and noise analysis. (a) Raw SPR phase signal in deionized water. The solid line represents the typical curve, while the light-colored lines show results from repeated experiments. (b) Denoised phase data in deionized water. The solid line represents the typical curve, while the light-colored lines show results from repeated experiments. (c) Raw SPR phase signal in dilute solution of different concentrations. (d) Denoised phase data in dilute solution of different concentrations. (e) Phase error comparison between raw and denoised data across different concentrations.

Figure 8

Protein antigen–antibody interaction curves between goat anti-rabbit IgG and rabbit IgG. (a) Denoised SPR curves and the fitted curves of the whole ROI in the proposed SPR system and denoised frame. (b) Associating and dissociating curves measured by Biacore 8 K. (c) Denoised SPR curves and the fitted curves of each ROI in the proposed SPR system and denoised frame. (d) ROIs are shown in the phase image and prism. (e) Denoised SPR phase signals of a single ROI across multiple tests in deionized water. The solid line represents the typical curve, while the light-colored lines show results from repeated experiments.