First Insights into Macromolecular Components Analyses of an Insect Meal Using Hyperspectral Imaging

Comments 1: Hyperspectral imaging (400 - 1000 nm) was used to pre - characterize the nutrition of Hermetia illucens (BSF), focusing on moisture and crude fat. Results were compared with those of traditional Type 65 wheat flour. The research design is rational and the methodology innovative. The findings have application significance for feed - industry quality control. The paper has a complete structure and rich experimental data. However, some parts need more supplementation and refinement to boost scientific rigor and readability. I recommend publishing it after a revision.

Response 1: We sincerely appreciate the reviewer’s positive feedback. We have carefully addressed all the points raised by the reviewer and revised the manuscript accordingly.

Comments 2:

1) It is suggested to add the latest literature in this field to enhance the research background of this paper. For example, Environmental Research, 2023, 232: 116389.；Anal. Chem. 2025, 97, 5, 2952–2962；Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2023, 284: 121785.

Response 2:

Thank you for your suggestions regarding the literature review. We have taken them into account, and some of the suggested references have been included, along with other relevant sources. As a result, the introduction has been improved, and the following additions have been made (Lines 92-96):

“Zhou et al. (2023) used HSI with effective band range 387-1035 nm to identify honey varieties and the adulteration of single honey varieties with different mass fractions of fructose syrup; to detect peanuts mildew (Zou et al., 2022); and also for identification of adulterated safflower seed oil (Zou et al., 2021).”

Comments 3:

2) In the experiment, only black soldier fly powder and wheat flour from a single batch were utilized. It is advisable to incorporate samples from diverse batches or sources to validate the general applicability of the method.

Response 3:

We appreciate your insightful suggestion. As this was a preliminary study, we used a single batch of both wheat flour and black soldier fly powder to ensure consistency and limit variability in our initial analyses. All experiments were performed in triplicate to strengthen the reliability of the results. This study aimed to define a methodology in which the use of HSI, within a limited spectral range, allowed the detection of selected components and their differentiation in samples of different nature, which we successfully demonstrated. However, we acknowledge that incorporating multiple batches would be valuable for assessing the broader applicability of our findings. Therefore, in future studies, to develop a robust quality control methodology, batch-to-batch variability will be considered.

Comments 4:

3) To ensure data consistency, details on the hyperspectral camera's specific model and light - source stability control methods, such as halogen - source pre - heating time and light - intensity calibration procedures, must be provided

Response 4:

Thank you for pointing out this lack of information. We have added a detailed description of the technical specifications of the camera and calibration procedure, according to the manufacturer, and type of illumination and heating time, according to previous experiments (Section 2.3 of the manuscript).

Comments 5:

4) In the Results and discussion part, compare the fat and moisture ranges in Hermetia illucens powder from literature (e.g., Lu et al. reported 20 - 40% fat in 2022). Analyze if the 28% fat content in this study meets expectations and discuss factors like larval growth stage and drying process

Response 5:

Thank you for your valuable suggestion. To address your comment, we compared the fat content observed in this study (28%) with the ranges reported in the literature, such as the 20-40% fat content mentioned by Lu et al. (2022). The fat content of 28% in our BSF meal is consistent with the lower end of this range, which is expected, considering that the sample provided by Entogreen is defatted. Several factors can influence the fat content of BSF powder, including the larvae’s growth stage and the drying process. In our case, the defatting process may have reduced the fat content compared to other studies, and differences in the larval growth stage and drying conditions could also account for variations in fat content. These factors highlight the need for further studies to explore how different growth stages and drying methods affect the nutritional composition of BSF.

In the Results and Discussion section (Lines 259-263), we have added the following:

“The BSF meal sample provided by Entogreen is defatted, so its fat content may vary when compared to other studies in literature where whole meal were evaluated. However, another factor that may contribute to these differences in fat content is the substrates used to feed the larvae (Tognocchi et al., 2024). According to Gutiérrez et al. (2020), the effect of diet is closely related to the nutrient composition of the insect feeding substrates.”.

Comments 6:

5) The interpretation of spectral similarity analysis results (SAM and Pearson coefficient) needs further depth. Specifically, the significance of the 0.664 SAM value should be determined. For greater rigor, add statistical tests like p - value calculation or compare with another research.

Response 6:

Thank you for pointing this out. The p-values for each metric were computed using a permutation test procedure described in Supplementary Material on the context of the study of the illumination source. This procedure adds more rigor to the results, as suggested.

Comments 7:

6) Optimize some figures and tables. In Figure 1b, clearly label key components like the camera, light source, and sample stage in the "experimental setup". For Table 1, break down "other compounds" into specific components such as crude protein and ash. If not, define "other compounds" clearly.

Response 7:

To enhance clarity and provide more detailed information, Figure 1 has been simplified and labeled accordingly. Additionally, new figures have been included to further illustrate the procedures, such as the calibration verification and the associated experimental apparatus (now presented as Figure 2), as well as the illumination stability study (included in Supplementary Material). Regarding Table 1, the term "Other compounds" has been clarified in the manuscript, specifying that it may correspond to protein, fiber, high-quality amino acids, and minerals. This revision provides a clearer understanding of the composition and ensures better interpretation of the data.

Comments 1: In this contribution, Liliana G. Fidalgo et al. have proposed an innovative approach to perform a nutritional precharacterization of Hermetia illucens (Black Soldier Fly - BSF) larvae meal, using hyperspectral images in the range 400-1000 nm, with a spectral resolution of 7 nm and a spatial sampling of 512 pixels, and correlate them to traditional chemical analysis methods. The chemical results of BSF meal indicated 7.2% ± 0.05% (w/w) and 28.15%±0.15% (w/w) in moisture and total fat content, respectively. Comparative analysis with wheat flour samples revealed that the hyperspectral imaging distinctly demonstrated a lower total fat content in BSF larvae meal. This study holds significant academic value and practical implications for the characterization and analysis of the distribution and content of nutrients in food and feed products. The topic is very interesting, but some improvement is needed before accepted in J. of Applied Sciences.

Response 1: We sincerely appreciate the reviewer’s positive feedback. We have carefully addressed all the points raised by the reviewer and revised the manuscript accordingly.

Comments 2: The article should provide more detailed results, discussing the reasons behind the lower fat content in BSF larvae meal compared to wheat flour, as well as analyzing how the fat content changes throughout the growth cycle of BSF larvae.

Response 2:

Thank you for your valuable comment. In response, we have added more detailed information regarding the fat content in the BSF larvae meal and its comparison with wheat flour (Lines 263-269):

“Normally, the results found in the literature indicate that BSF meal contains 50% crude protein, 35% lipids, and have a quite interesting amino acid profile (Elwert et al., 2010). However, the sample used in this study does not represent the standard BSF or Entogreen production pad, where it is only partially degreased. This process involves pressing several dehydrated larvae to remove a large percentage of fat, which justifies the lower fat percentage observed compared to other studies.”

Comments 3: The data collection section is typically detailed; it is recommended that the authors provide a more comprehensive presentation of the sample preparation process, testing methods, and repeatability aspects.

Response 3:

Thank you for this remark. In fact, Figure 5 presents raw, non-processed spectra, with the mean spectra highlighted in cyan. These mean spectra, extracted from the respective ROIs, serve as the basis for the subsequent steps of the processing workflow.

Comments 4: It is suggested that the authors include pre-processing results in the manuscript and conduct a comparative analysis with the post-processed results.

Response 4:  

Thank you for pointing this out. The entire manuscript has been carefully reviewed to correct any grammatical and typographical errors.

Comments 5: Carefully proofread the manuscript to correct any grammatical and typographical errors.

Response 5:

Thank you for your suggestion. We have carefully proofread the manuscript to correct any grammatical and typographical errors.

Comments 1: The study explores the non-invasive nutritional analysis of feed using hyperspectral imaging, an innovative biotechnology approach. By capturing images across multiple wavelengths, this technique identifies unique spectral patterns associated with nutritional components such as total fat and moisture. The findings highlight the potential of this technique for broader feed characterization, including protein content assessment, reinforcing its role in advanced nutritional analysis., but some improvement is needed before accepted in J. of Applied Sciences.

Response 1: We sincerely appreciate the reviewer’s positive feedback. We have carefully addressed all the points raised by the reviewer and revised the manuscript accordingly.

Comments 2: The work is relatively well organised. The introductory part provides the necessary information that later explains the necessity of the research. The methods are well described, sliek and tables are structured and follow the text, and in this sense there is no need for corrections. Only chapter 2.3 would be good. As it contains the results and not the methodology, it should be included in the chapter where the results are presented, but renamed as preliminary research.

Response 2:

Thank you for this suggestion. Section 2.3 has been revised to focus solely on the experimental apparatus, equipment, and procedures during data acquisition, ensuring it remains part of the methodology. This section now contains essential details regarding equipment specifications and calibration checks. Additionally, Section 2.4 has been structured to detail the processing performed by the acquisition system, including reflectance computation, spectral sampling from the datacubes, and the mathematical methods used for spectral analysis. The results derived from these methods are presented and discussed in Section 3.

Comments 3: In a part of the text that refers to: "The similarity of the two spectra can be evaluated using different metrics, namely the Pearson correlation coefficient and the spectral angle mapper (SAM) [24], which calculates the angle between the two SNV-transformed spectra when they are considered as vectors in a multidimensional space, where the number of dimensions is equal to the number of bands in the hypercube".

It is not clear whether these are recommendations or whether measurements have actually been made. If so, they should be described in the methods.

Response 3:

Thank you for this comment. Both the Pearson correlation coefficient and the spectral angle mapper (SAM) metrics were indeed calculated. The calculation process is now described in Section 3, “Mean Spectrum Processing,” as part of the data analysis workflow. Additionally, in Section 3, "Results and Discussion," we present the values of these metrics, along with the corresponding p-value calculations, to provide statistical rigor to the results.

Comments 1: This manuscript investigates the application of hyperspectral imaging (HSI) for the nutritional characterization of Hermetia illucens (Black Soldier Fly) larvae meal. The study aims to correlate HSI spectral data with traditional chemical analysis methods, focusing on moisture and fat content. Given the growing interest in alternative protein sources and non-destructive analytical techniques, this research is highly relevant to food and feed sciences.

Response 1: We sincerely appreciate the reviewer’s positive feedback.

Comments 2: The manuscript offers valuable insights and has strong potential for publication. However, some revisions are necessary to enhance its clarity, coherence, and overall impact. I kindly encourage the authors to consider the following suggestions, which will help strengthen the manuscript and improve its chances of acceptance:

Response 2: We have carefully addressed all the points raised by the reviewer and revised the manuscript accordingly.

Comments 3: Methodological Considerations: The authors should clarify the rationale behind selecting the spectral range of 400–1000 nm. Some studies suggest that extending the range to include the near-infrared (NIR, >1000 nm) region could improve the reliability of the results. For instance, Li et al. [Front Plant Sci. 2023;14:1275004. doi: 10.3389/fpls.2023.1275004] discuss the benefits of this broader range in similar applications.

Response 3:

Thank you for this very important comment. The limited spectral range of the study was imposed by the equipment available, namely, a HSI camera in the range VNIR (400 – 1000 nm). This preliminary study had a twofold objective: to establish a methodology to compare analytical and HSI data and detect common features even with a limited spectral range, and to develop the HSI processing software that could be used in the future with data cubes with an extended spectral range. The study shows that possibility and provides the methodology to extend the data range in the near future with new equipment. In the conclusions we already highlight the importance of extending the spectral range but we would like to thank the reviewer for the information related to a relevant reference for this topic, which we have appreciated and included (Lines 311-312):

“Future work will address this spectral range extension and the continuous upgrade of the software to include different HSI processing tools.”.

Comments 4: Discussion Depth: The discussion should provide a more detailed analysis of the spectral differences between wheat flour and insect meal. Emphasizing the implications of the obtained spectral signatures would enhance the findings' scientific value.

Response 4:

Thank you for this suggestion. The present study is limited by the spectral range available at the moment, and as such, the discussion primarily focuses on the spectral features that were identifiable using the current methodology. We acknowledge that the spectral differences between wheat flour and insect meal could be explored further with more advanced algorithms and extended spectral ranges. The current validation serves as a foundation for future work, which will include a broader spectral range, enabling a more comprehensive analysis. Despite these limitations, we highlight the potential industrial applications of our findings, particularly in real-time and non-destructive food analysis.

Comments 5: Conciseness: Certain sections contain redundant information, particularly the introduction and methodology. Streamlining these sections would improve readability and overall manuscript structure.

Response 5:

Thank you for your comment regarding the conciseness of the manuscript. We have reviewed the introduction and methodology sections and made efforts to streamline them by removing redundant information.

Comments 6: Literature Review: Some references are outdated or irrelevant to the study. The literature review should be updated to include more recent studies on HSI applications in food sciences to ensure a comprehensive contextual framework.

Response 6:

Thank you for your suggestion regarding the literature review. We appreciate your input and have made efforts to ensure the review remains comprehensive and up-to-date. While the majority of references used are from the past six years to reflect recent developments, some older references were included because they are fundamental to the understanding of the methods and concepts applied in this study. For example, publications like Barnes (1989), Savitzky & Golay (1964), and Pope (1997) are still relevant as they describe key methodologies that continue to be cited in current literature. These references are crucial for supporting foundational claims, particularly with regard to the O-H harmonics. We believe referencing the original authors provides necessary historical context for the methodologies used, while also ensuring the validity of our approach.

Comments 7: Addressing these points will significantly strengthen the manuscript, increasing its clarity and impact within the scientific community.

Response 7: Thank you for your valuable feedback, which has contributed to refining our work.

Comments 1: In this manuscript, Silva et al. attempt to explore the application of near-infrared hyperspectral imaging (NIR-HSI) for the non-destructive nutritional characterization of feed components—specifically, Black Soldier Fly (BSF) larvae meal compared to wheat flour. This manuscript is methodologically rigorous, with detailed data acquisition and pre-processing procedures, combining both traditional chemical analyses and in-depth mining of high-dimensional spectral data using image-processing techniques, reflecting a high level of experimental and data-processing sophistication.

Response 1: We sincerely appreciate the reviewer’s positive feedback.

Comments 2: Major comments:

1.      Although it was mentioned in the manuscript that the samples were all from the same batch, it is recommended that comparisons of multiple batches of samples be added in future studies to further validate the robustness and generalizability of the method.

Response 2:

Thank you for your valuable suggestion. As this was an initial study, we used a single batch to ensure consistency and reduce variability in our preliminary analyses. However, we recognize the importance of comparing multiple batches to further validate the robustness and generalizability of the method. This aspect will be considered in future studies as we advance in our research.

Comments 3:

2.      The current study focuses on water and fat components, and subsequent work could consider using the extended spectral range to conduct more in-depth investigations for other key nutrients such as protein, fiber, and minerals to enrich the dimensions of overall nutritional analysis.

Response 3:

We appreciate the reviewer’s valuable suggestion regarding the potential for further investigations using an extended spectral range to analyze additional key nutrients such as protein, fiber, and minerals. While our current study focused on moisture and crude fat, we recognize that expanding the spectral range, such as into the near-infrared region, could offer important insights into other nutritional components and enrich the overall nutritional analysis of Hermetia illucens.

This is indeed an important direction for future research. At the time of this study, our laboratory did not have the capacity to extend the spectral range. However, our primary goal was to develop a methodology and spectral range-agnostic software that can be applied to future data acquisitions with new equipment, facilitating more in-depth analysis of a broader range of nutrients. We appreciate your suggestion and will certainly consider this approach in future studies.

Comments 4:

3.      For the preprocessing part of the HSI data, although standard normal transform and Savitzky-Golay filtering were used, it is recommended to discuss other possible methods of noise reduction and data correction with a view to obtaining better spectral resolution and analytical accuracy.

Response 4:

Thank you for this important suggestion. As this was a preliminary study, we focused on a standard approach to spectrum processing, utilizing standard normal variate (SNV) transformation and Savitzky-Golay filtering. We acknowledge that other noise reduction and data correction methods could further enhance spectral resolution and analytical accuracy. In future studies, we plan to explore and compare additional preprocessing techniques to optimize data quality and improve the robustness of our methodology.

Comments 5:

4.      The manuscript involves several key images (e.g., ROI selection, mean spectra, SG second-order derivative plots, etc.), and it is recommended to further optimize the quality of the graphs in the final version (especially Fig. 1 a), with detailed labels on the legends and descriptions, to help readers more intuitively understand the data processing and result comparison.

Response 5:

Thank you for this suggestion. The figures have been enhanced to improve clarity and readability. Specifically, Figure 1a was simplified and labeled for better comprehension. Additional figures related to the calibration procedure were included to provide more context. Furthermore, all graphs, labels, and legends have been refined to align with the journal's template. To ensure high-quality visualization, all figures were generated using Matplotlib at 600 dpi.

Comments 6:

5.      The interactions between different feed components and specific use cases of hyperspectral technology in real-time quality control can be further discussed to enhance the applied translational value and practical guidance of the paper.

Response 6:

Thank you for your insightful comment. In response to your suggestion, we have added a statement in the conclusions emphasizing the importance of exploring interactions between different feed components and the specific applications of hyperspectral technology in real-time quality control. We acknowledge that this is a growing field with significant implications for both health and food industries, and we consider it an important avenue for future research (Lines 313-317):

“In the future, the interactions between different feed components and the specific applications of hyperspectral technology in real-time quality control can be further explored to enhance their translational value. This topic has been growing in recent years and will play a crucial role in establishing important correlations for the health and food industries.”