Detection of microplastics stress on rice seedling by visible/near-infrared hyperspectral imaging and synchrotron radiation Fourier transform infrared microspectroscopy

Abstract

Introduction: Microplastics (MPs), as emerging environmental contaminants, pose a significant threat to global food security. In order to rapidly screen and diagnosis rice seedling under MPs stress at an early stage, it is essential to develop efficient and non-destructive detection methods.

Methods: In this study, rice seedlings exposed to different concentrations (0, 10, and 100 mg/L) of polyethylene terephthalate (PET), polystyrene (PS), and polyvinyl chloride (PVC) MPs stress were constructed. Two complementary spectroscopic techniques, visible/near-infrared hyperspectral imaging (VNIR-HSI) and synchrotron radiation-based Fourier Transform Infrared spectroscopy (SR-FTIR), were employed to capture the biochemical changes of leaf organic molecules.

Results: The spectral information of rice seedlings under MPs stress was obtained by using VNIR-HSI, and the low-dimensional clustering distribution analysis of the original spectra was conducted. An improved SE-LSTM full-spectral detection model was proposed, and the detection accuracy rate was greater than 93.88%. Characteristic wavelengths were extracted to build a simplified detection model, and the SHapley Additive exPlanations (SHAP) framework was applied to interpret the model by identifying the bands associated with chlorophyll, carotenoids, water content, and cellulose. Meanwhile, SR-FTIR spectroscopy was used to investigate compositional changes in both leaf lamina and veins, and two-dimensional correlation spectroscopy (2DCOS) was employed to reveal the sequential interactions among molecular components.

Discussion: In conclusion, the combination of spectral technology and deep learning to capture the physiological and biochemical reactions of leaves could provide a rapid and interpretable method for detecting rice seedlings under MPs stress. This method could provide a solution for the early detection of external stress on other crops.

Keywords: deep learning; microplastics; rice seedlings; synchrotron radiation-based Fourier transform infrared spectroscopy; visible/near-infrared hyperspectral imaging.

Figure 1

The extraction process of macro VNIR spectra. (a) Original pseudo-color image; (b) spectra of the background and leaves; (c) Grayscale images at 655nm; (d) Mask image; (e) Apply the original image to the mask image to remove the background; and (f) Mean spectra of each leaf.

Figure 2

The framework of SE-LSTM model. (a) Overall structure of the model; (b) SE block attention mechanism structure; (c) LSTM cell structure.

Figure 3

Original and mean spectra of leaves of rice seedling under PET (a, d), PS (b, e), PCV (c, f) MPs stress.

Figure 4

The scatter and loading plot of PET (a–c), PS (d–f) and PVC (g–i) stress on rice seedlings.

Figure 5

Accuracy and loss value of SE-LSTM model with different hidden layer and sizes.

Figure 6

Accuracy (a–c) and loss value (d–f) of the SE-LSTM model with different learning rates (0.001, 0.005, 0.01, 0.05, 0.1) and optimizers (SGD: a,d; Adam: b,e; RMSprop: c,f).

Figure 7

Accuracy (a) and overfitting index (b) of SE-LSTM model with different learning rates and loss function.

Figure 8

Selection of characteristic wavelengths for rice seedling stressed under PET, PS and PVC.

Figure 9

Performance for the classification models based on characteristic wavelengths.

Figure 10

Importance distribution of characteristic wavelength based on SHAP values for (a) the overall dataset, (b) the control group, (c) the 10 mg/L group, and (d) the 100 mg/L group.

Figure 11

One dimensional average spectra and PCA loading plots of leaf veins (a, b) and mesophyll (c, d).

Figure 12

The synchronous (a, c) and asynchronous (b, d) 2DCOS maps generated from the SR-FTIR spectra of leaf veins in the functional group region and fingerprint region.

Figure 13

The synchronous (a, c) and asynchronous (b, d) 2DCOS maps generated from the SR-FTIR spectra of leaf mesophyll in the functional group region and fingerprint region.