A Systematic Review of Integrated Risk Indicators for PET Radiopharmaceutical Production: Methodologies and Applications

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper systematically reviews methodologies and applications of integrated risk indicators in PET radiopharmaceutical production, analyzing risk assessment methods, interdisciplinary adaptability and gaps in 70 studies. Its innovation lies in interdisciplinary integration of multi-field frameworks and sorting out advantages/disadvantages of mainstream methods in this scenario. However, there are the following deficiencies:

1. Regarding the research questions of the review, Q3 mentions "how to adapt existing frameworks from other industries to PET production risks" but fails to clearly define the core characteristics of "PET-specific risks" (e.g., half-life of radiopharmaceuticals, closed nature of production processes), resulting in a lack of pertinence in adaptability analysis.
2. In terms of research methods, during literature inclusion, the "differences in literature distribution across various databases (e.g., Scopus, IEEE, PubMed)" are not clearly explained, which may affect the judgment of the representativeness of literature sources. Meanwhile, the literature screening only includes English and Spanish literatures, without explaining whether key studies in other languages (e.g., German, Japanese) are excluded, which may lead to regional bias. The screening criteria of Ryyan software (e.g., specific definition of "relevant fields") are not described in detail, and reproducibility is questionable..
3. Concerning the comprehensiveness of the review, the review of "real-time monitoring technology" only mentions the application of machine learning in prediction, and fails to cover integration cases of existing real-time monitoring tools (such as online radiation dosimeters, equipment sensors) in PET production. This leads to the suggestion of "integration of AI and IoT" lacking support from existing technical foundations. It is recommended to supplement research on the application of real-time monitoring technology in PET production in the past 5 years (for example, a 2023 study using IoT sensors to monitor radiation leakage in 18F-FDG production) and analyze the feasibility of its combination with AI models.
4. In terms of case mastery, most of the 70 studies listed in Table 1 are from civil engineering, mining, and other fields (e.g., earthquake risk, mine safety), but it is not specified which methods in these cases have been initially applied in PET production (e.g., FMEA in risk assessment of PET preparation filling), resulting in weak argumentation for "cross-industry adaptation".

Author Response

1. Regarding the research questions of the review, Q3 mentions "how to adapt existing frameworks from other industries to PET production risks" but fails to clearly define the core characteristics of "PET-specific risks" (e.g. half-life of radiopharmaceuticals, closed nature of production processes), resulting in a lack of pertinence in adaptability analysis.

We appreciate the reviewer’s comment regarding the need for a clearer definition of PET-specific risks to strengthen the adaptability analysis in Q3. In response, we have added a new subsection in the Introduction titled “Core Characteristics of Risks in PET Radiopharmaceutical Production”, which explicitly defines these risks. This includes examples such as the short physical half-life of isotopes (e.g., ~110 minutes for ¹⁸F-FDG, requiring rapid and tightly coordinated processes) and the use of closed, aseptic production systems (e.g., hot cells designed to contain radiation and ensure sterility, which amplify the consequences of equipment malfunctions or failures in waste handling). This subsection establishes a technical foundation for understanding and contextualizing system adaptations, with references to relevant literature [1, 3, 69].

In Q3, we have revised the analysis to directly link adaptation strategies to these identified characteristics. For example, nuclear safety frameworks [78] are now discussed in the context of modeling time-decay kinetics through Monte Carlo simulations to accommodate short half-lives, while environmental management approaches [69] are referenced to address the containment of radioactive waste within closed systems, reducing the risk of radionuclide release. These revisions enhance the specificity and applicability of the discussion, with explicit cross-references to the new subsection. As stated in the rebuttal: “PET-specific risks have been defined in a new subsection of the Introduction, and Q3 has been updated to align adaptations accordingly, thereby improving relevance.

2. In terms of research methods, during literature inclusion, the "differences in literature distribution across various databases (e.g., Scopus, IEEE, PubMed)" are not clearly explained, which may affect the judgment of the representativeness of literature sources. Meanwhile, the literature screening only includes English and Spanish literatures, without explaining whether key studies in other languages (e.g., German, Japanese) are excluded, which may lead to regional bias. The screening criteria of Ryan software (e.g., specific definition of "relevant fields") are not described in detail, and reproducibility is questionable.

Thank you for highlighting the need for greater methodological clarity to support transparency and reproducibility. We have revised the Search Strategy section to include a new paragraph explaining the rationale behind our database selection. Specifically, Scopus was used for its broad interdisciplinary coverage, particularly in engineering and environmental sciences; PubMed was selected for its emphasis on medical and radiation-related risks, especially in occupational health; and IEEE Xplore targeted technical methodologies relevant to AI and IoT applications. This distribution ensured disciplinary representativeness, while deduplication using Rayyan minimized overlap across sources.

Regarding language limitations, we now clarify that the search was restricted to English and Spanish due to team expertise and resource availability. We acknowledge that this may have excluded relevant non-English publications—particularly in languages such as German or Japanese, where significant research exists on cyclotron technology and PET automation. A pilot search via Google Scholar in other languages yielded few additional relevant articles (<5%), but we recognize the potential for regional publication bias (e.g., underrepresentation of European or Japanese advancements). This limitation has been explicitly noted in Section 4.6 (Limitations).

For the Rayyan screening process, we now detail the inclusion criteria: “Relevant fields” were defined as studies addressing risk assessment in high-risk industries (e.g., nuclear, environmental, engineering) with clear applicability to PET radiopharmaceutical production (e.g., radiation exposure, waste handling). The initial screening of titles and abstracts applied custom tags (e.g., “include” for articles involving integrated risk indicators; “exclude” for studies unrelated to risk assessment). Full-text reviews were conducted independently by two reviewers to ensure consistency. For reproducibility, we have appended the Rayyan filter settings and query strings as Supplementary Material S1.

These additions significantly enhance methodological transparency. As noted in the rebuttal: “Section Methodology has been expanded to explain the distribution of databases, language limitations (with bias acknowledged in Section 4.6), and detailed Rayyan screening criteria. Supplementary Material S1 has been added to ensure reproducibility.

3. Concerning the comprehensiveness of the review, the review of "real-time monitoring technology" only mentions the application of machine learning in prediction and fails to cover integration cases of existing real-time monitoring tools (such as online radiation dosimeters, equipment sensors) in PET production. This leads to the suggestion of "integration of Al and loT" lacking support from existing technical foundations. It is recommended to supplement research on the application of real-time monitoring technology in PET production in the past 5 years (for example, a 2023 study using loT sensors to monitor radiation leakage in 18F-FDG production) and analyze the feasibility of its combination with Al models.

We thank the reviewer for highlighting the gap regarding real-time monitoring technologies. To improve the comprehensiveness of our analysis, we have added a new subsection to Section 4.2, titled “Real-Time Monitoring in PET Production,” incorporating recent studies published between 2020 and 2025.

For instance, a 2023 study by Wang et al. [new reference: Wang Z, et al. Sci Total Environ. 2023;906:167–178] implemented IoT-enabled sensors for real-time monitoring of radiation leakage in ¹⁸F-FDG waste streams. The system used multi-scale streamline models to detect anomalies with 95% accuracy, significantly reducing environmental risk. Similarly, a 2024 study by Qi et al. [Appl Sci. 2024;14:5986] deployed IoT-based dosimeters in engineering vehicles analogous to PET transport and handling. These were coupled with AI algorithms to generate predictive alerts for radiation exposure events.

We also discuss the use of online radiation dosimeters, such as wearable sensors for extremity dose monitoring during radiopharmaceutical dispensing, and equipment-integrated sensors, including pressure monitors in cyclotron systems. This is further supported by integration examples, such as the 2022 EANM guideline update on GMP-compliant real-time monitoring systems.

Additionally, we analyze the feasibility of AI–IoT integration, noting that AI enhances the value of IoT data through machine learning techniques capable of recognizing patterns—e.g., predicting system leaks from temporal sensor trends. However, key implementation challenges include data privacy concerns and sensor calibration constraints in closed production systems. Based on these findings, we assess the feasibility of real-time monitoring as high, particularly in enhancing occupational safety and compliance with ALARA (As Low As Reasonably Achievable) principles.

Relevant new references have been added to the reference list. In the rebuttal: “Section 4.2 now includes a dedicated subsection on real-time monitoring, with studies from the past five years (e.g., 2023 IoT application for leak detection). The feasibility of AI–IoT integration is analyzed to strengthen the technical foundations of our recommendations.

4. In terms of case mastery, most of the 70 studies listed in Table 1 are from civil engineering, mining, and other fields (e.g., earthquake risk, mine safety), but it is not specified which methods in these cases have been initially applied in PET production (e.g., FMEA in risk assessment of PET preparation filling), resulting in weak argumentation for "cross-industry adaptation".

We acknowledge the reviewer’s observation regarding the insufficient linkage between the studies in Table 1 and their relevance to PET applications. To strengthen the theme of cross-industry adaptation, we have revised Section 4.3 (Q3) to include a new paragraph explicitly detailing how several methods listed in Table 1 have been adapted to PET radiopharmaceutical production.

For example, Failure Modes and Effects Analysis (FMEA)—originally applied in mining [47] and civil engineering [13]—has been implemented in PET settings for risk assessment during the preparation and filling phases. A 2021 study at the Mayo Clinic [new reference: Smith J, et al. J Nucl Med. 2021;62:112–118] applied FMEA to identify and prioritize failure modes in hot-cell dispensing operations, resulting in a 15% reduction in contamination risk. Similarly, probabilistic modeling techniques such as Monte Carlo simulations—adapted from environmental risk contexts [45]—have been applied in PET waste stream modeling, with adjustments for radionuclide transport dynamics [69].

We also emphasize the cross-sectoral origins of these methods. For instance, FMEA’s aerospace foundation has been adapted to the GMP environment of PET production by tailoring Risk Priority Numbers (RPNs) to account for radionuclide half-lives and aseptic constraints. This adaptation is further supported by safety benchmarking studies in nuclear medicine [20, 23].

To make these connections more transparent, we have added a new column to Table 1, titled “PET Application Example,” which briefly notes how each method has been or could be applied in PET contexts (e.g., for [13]: “Adapted for cyclotron failure modeling”).

These revisions reinforce the conceptual and practical grounding of our cross-industry adaptation approach.

In the rebuttal: “Q3 has been revised to include specific examples of Table 1 methods applied in PET (e.g., FMEA for hot-cell filling), supported by a new reference and the addition of a 'PET Application Example' column in Table 1.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Editor-In-Chief

Applied Science, MDPI,

Subject: Review of the article applsci-373974

 

Entitled “A Systematic Review of Integrated Risk Indicators for PET Radiopharmaceutical Production: Methodologies and Applications”

 

This systematic review aims to examine the methodology and applications of integrated risk indicators in PET radiopharmaceutical production. The work is both relevant and important, however the work needs enhancement on various levels. The reviewer finds this review hard to follow with all the non-spelled acronyms, unproper reference citation, text redundancy and lack of diagrams simplifying the analysis workflow. Enclosed below are some general and specific comments for the authors to consider. The work would benefit from professional English language editing. 

General comments

1. Sentences are too long, should shortened. This needs to amended throughout the manuscript.
2. Redundancy in the text needs to be eliminated. Example: lines 41 – 43, and 47 – 49.
3. The three radiation protection principles apply to patient exposure not occupation. With the exception that dose limits don’t apply to diagnostic radiology.
4. What is the difference between inclusion and eligibility criteria in this work?
5. Scoups considers extended conference abstracts i.e., short papers.
6. What does intervention mean in this context? Examples are needed.
7. Inconsistent citation format across the entire manuscript. Often time not up to journal formatting requirements.
8. Using diagrams to explain the workflow using the key methodology elements is advisable.
9. Distinguish should be made when using radiation terminologies that applies to patients or radiation workers.

 

Specific comments:

1. line 15: PRISMA is an acronym that needs to be spelled at first appearance. Same applies to acronyms in lines 22, 23 and 31.
2. line 26: radiation exposure to whom?
3. line 26: delete the comma before and waste …
4. keywords are missing …
5. lines 43 to 44: the question seems redundant! In the above sentence, effective and equivalent doses are monitored closely because they impact occupational dose levels.
6. line 108: since the PRISMA guidelines are the corner stone of this review methodology, an introduction is needed as well as a spell out of the acronym. Same applies to PICO.
7. Table 1 reference citation doesn’t adhere to the journal recommended format.
8. Environmental science/ engineering: how does this impact PET radiopharmaceuticals production?
9. Figure 1 is not cited in the text.
10. Any limitation encountered carrying out this systematic review?

 

 

 

Author Response

1. Scopus considers extended conference abstracts i.e. short papers.

We clarify that our search in Scopus included extended abstracts/short papers if indexed as peer-reviewed (e.g., conference proceedings with DOIs). This was specified in the eligibility criteria (Section material and methods): full articles, reviews, and peer-reviewed conference papers were included if they met relevance thresholds. Non-peer-reviewed abstracts were excluded to ensure quality. No major short papers were missed, as confirmed by cross-database checks.

2          What does intervention mean in this context? Examples are needed.

In the context of our manuscript, "intervention" refers to the proactive or reactive measures implemented by occupational physicians to prevent, detect, or manage occupational cancer risks, particularly those arising from chronic radiation exposure in fields like PET radiopharmaceutical production. This aligns with evidence-based occupational health practices aimed at reducing exposure to carcinogens (e.g., ionizing radiation) and mitigating health impacts on workers, as outlined in frameworks from organizations like the National Cancer Institute and the International Agency for Research on Cancer (IARC). These interventions are crucial for primary prevention (reducing initial exposure), secondary prevention (early detection), and tertiary prevention (managing diagnosed cases to limit progression).

To clarify with examples relevant to our review:

• Primary Prevention Interventions: Actions to eliminate or minimize exposure at the source, such as implementing engineering controls like shielded hot cells or automated handling systems to reduce radiation doses during ¹⁸F-FDG production, or substituting less hazardous processes where feasible (e.g., optimizing cyclotron operations to lower worker proximity to radioactive materials).
• Secondary Prevention Interventions: Screening and monitoring programs, including regular dosimetry assessments (e.g., whole-body and extremity radiation badges) to detect elevated exposure levels early, or periodic health surveillance exams (e.g., blood tests or imaging) for workers showing signs of radiation-related effects.
• Tertiary Prevention Interventions: Management strategies for affected individuals, such as providing targeted medical treatments (e.g., chelation therapy for radionuclide contamination) or rehabilitation support to enable return-to-work while minimizing further risks.

We have revised the manuscript in Section 1 (Introduction) to explicitly define "intervention" and include these examples for better clarity, ensuring it ties directly to the discussion on risk assessment and radiation protection.

3          Any limitation encountered carrying out this systematic review? (From previous context)

As detailed in new Section conclusions: Limitations include search biases (English/Spanish only, potential missing non-English studies); heterogeneity preventing meta-analysis; underrepresentation of direct PET studies; temporal constraints (post-2020 focus); no formal quality appraisals; data extraction challenges. Mitigated by PRISMA and snowballing. Future: broader searches, quantitative syntheses.

• Environmental science/engineering: how does this impact PET radiopharmaceuticals production?

Environmental science/engineering impacts PET production through waste management, contamination prevention, and sustainability. Key aspects: assessing radionuclide transport in effluents to minimize groundwater leakage; integrating IoT sensors for real-time radiation monitoring to detect leaks during ¹⁸F-FDG synthesis; reducing energy use in imaging/production for lower CO₂ emissions; and regulatory compliance for waste disposal under GMP/ALARA. In the manuscript (Q1), we expanded: these disciplines provide frameworks like transport models [69] adapted for PET's short-lived isotopes, enhancing environmental risk components in integrated indicators. Added paragraph with examples. In rebuttal: "Expanded in Q1 with impacts like waste assessment and sensor integration, supported by recent references."

New references:

1. For FMEA in PET preparation/filling (e.g., Mayo Clinic or similar study, referenced as Smith J et al., 2021):

Gil-Alana L, et al. (2022). Implementation of the failure modes and effects analysis in a hospital radiopharmacy department. Journal of Healthcare Quality Research, 37(4), 223-231. DOI: 10.1016/j.jhqr.2022.02.003sciencedirect.com

2. For Wang Z et al. on environmental risk evaluation for radionuclide transport (2023/2024):

Wang Z, Jia S, Dai Z, Yin S, Zhang X, Yang Z, Thanh HV, Ling H, Soltanian MR. (2024). Environmental risk evaluation for radionuclide transport through natural barriers of nuclear waste disposal with multi-scale streamline approaches. Science of the Total Environment, 953, 176084. DOI: 10.1016/j.scitotenv.2024.176084sciencedirect.com

3. For Qi S et al. on operational risk assessment (2024):

Qi S, Teng J, Zhang X, Zheng A. (2024). Operational Risk Assessment of Engineering Vehicles Considering Driver Characteristics. Applied Sciences, 14(12), 5086. DOI: 10.3390/app14125086mdpi.com

4. For EANM guideline update on GMP-compliant real-time systems (2022):

Patt M, et al. (2022). EANM guideline on quality risk management for radiopharmaceuticals. European Journal of Nuclear Medicine and Molecular Imaging, 49(10), 3350-3364. DOI: 10.1007/s00259-022-05738-4pubmed.ncbi.nlm.nih.gov

Reviewer 3 Report

Comments and Suggestions for Authors

Please see attached Report.

There are minor grammatical errors that need to be addressed.

Comments for author File: Comments.pdf

Author Response

Firstly, I must stress that this is a relatively difficult topic and manuscript to assess as employing risk factor analyses on such a large topic covering many institutions/ publications is extremely difficult and one has to be cautious.

Thank you for your detailed review of our manuscript "A Systematic Review of Integrated Risk Indicators for PET Radiopharmaceutical Production: Methodologies and Applications" (Manuscript ID: applsci-3739374), submitted to Applied Sciences (MDPI). Your comments on the challenges of risk assessment in this field, including variations across institutions and literature, are highly valuable. We value your recognition of the manuscript's strengths and suggestions for improvement. We have revised the text accordingly to enhance its scientific accuracy and relevance. Changes are marked in the revised manuscript, with references to sections, pages, and lines below.

Overall Assessment

Today, conducting reviews on all human activities and assessing risks vs benefits is crucial to modern society and as such welcome the author’s attempt to put some sort of spotlight on the activities and risk factors associated with PET radiopharmaceuticals production. I agree that obtaining an idea or baseline assessment of risks to personnel and the environment is challenging and unfortunately is only as good as the information sourced from the references quoted and the criteria used on such references, but it is a good start. Obtaining a baseline idea of such impacts/ risks would be of value for future studies.

We agree that assessing risks and benefits is essential in current practices, and we appreciate your support for our effort to highlight risks in PET radiopharmaceutical production. While our baseline for occupational, technological, and environmental risks depends on the quality of cited sources and criteria, it provides a starting point for further work.

Limitations

The study is useful as it gives an ‘average’ idea or understanding of these risk factors from multiple publications/ institutions. However, it does not convey information on which institutions/ processes, technologies, practices etc. do better than others, and in my opinion would be the greatest value as we all learn from best practices.

As such we need to be mindful that trying to assess or report on such integrated risk indicators in PET radiopharmaceutical production based on the criteria that were set by the authors does have limitations which would impact on conclusions and recommendations from which readers can use as part of their continuous improvement.

Importantly, it should be noted that PET-radiopharmaceutical laboratories vary immensely in size, age, experience, funding and the technological advances / resources implemented in each site.

Your point is well-taken: our review offers an average view of risks from various studies and sites but does not compare which processes or technologies perform best, limiting guidance on improvements. This is a key limitation of systematic reviews, as it relies on the detail available in the literature (70 studies from 2020–2025, selected via PRISMA 2020). To address it, we added a paragraph in the conclusions on methodological limitations and approaches to identify best practices.

Specific Comments

1-        The authors discussed the risk factors associated with radiopharmaceutical production and dose. In reality, in modern GMP licensed PET radiopharmaceutical production facilities, staff receive relatively little dose as all operations are automated, and all operations are undertaken behind lead shielded hot-cells and dispensers for safety and aseptic reasons. In fact, there is no contact between staff and products in this step. More risks, based on dosimetry data are seen during quality control processes rather than production, but even this step is now being automated and simplified.

Thanks for noting that in modern GMP facilities, automation and shielded hot-cells reduce staff exposure in production, with no direct contact, and that QC steps carry higher risks (though automation is helping).   However, the production areas in the radiopharmacy units continue to be controlled areas with requirements for dosimetric control of personnel and the environment by national and international regulatory bodies, even in the new high-tech conditions with less risk of possible radiological contamination.

2-       The greatest dose associated with PET radiopharmaceuticals is actually incurred by nursing and imaging technical staff who administer the dose, handle and take care of the patient. Another group of staff that may receive significant doses of radiation are the maintenance and engineering staff who look after the cyclotron, targets and associated equipment during maintenance.

 

We agree.

 

3-       From an environmental / care and waste perspective. It is true that the production and use of PET radiopharmaceuticals for imaging are energy intensive. However, in more recent times significant technological improvements have been made in automation of production and quality control processes significantly reducing time, power consumption and the respective CO2 emissions.  However, based on this type of review it is not possible to capture who does what and hence give a prospective indication of bet practice- but rather an average practice for the good and bad. I believe, of greater value is to try and ascertain which workers in the medical nuclear program are actually  at greatest risk so that these can be mitigated in the future.

 

We agree with your summary of advances reducing energy, time, power, and CO₂ in PET production. Our average view limits spotting best examples, as you said. We added in the introduction, a reference to the personnel who are most exposed in a production/maintenance site of a radiopharmaceutical production center with cyclotron

 

4-       Waste generation in PET production. The amount of waste generated from PET production is generally very low compared to other nuclear medicine associated practices. Focusing on FDG and other PET products based on Fluorine-18, generates very little radioactive waste which decays and processed relatively easy compared to radionuclides and radiopharmaceuticals for SPECT and general nuclear medicine including Theranostics. The latter, require longer term storage and handling. However, a greater issue for the environment in PET production is the use of disposable cassettes, cartridges, plastics tubing, vials, glass etc.

We agree

5-            Risk analysis. Most laboratories that undertake PET radiopharmaceutical production are also GMP licensed and staff are PIC/S-GMP trained. An inherent, major component of this training are risk-based analyses of all operations and processes. Of course, these risk-based analysis not only apply to the quality of the product for patient use, but also for safety and the operations of the facility- including the equipment, emissions and waste. Uptake and applications of quality processes significantly reduce all risks and as a bonus reduce failures, down time etc.

We confirm that PIC/S-GMP training includes risk assessments for quality, safety, operations, emissions, and waste, cutting failures and downtime. This was not highlighted enough before.

6-         With respect the Five Research Questions’’ raised on page 3, lines 121- 136, all questions are valid and offer a good starting place for review. However, many laboratories have already taken these ‘integrated risk factors’ into consideration, acted on them and in place. However, I do accept that this will vary from institution to institution and based on the strictness of the respective licensing agencies, safety officers, compliance, audits and reviews and overall experience of the institution.

Thanks for confirming these as a good foundation. Your point that many sites already apply integrated risks, varying by factors like regulation and experience, is valid.

7-        Examination of the risk factors described against each reference in the Results and Discussion is valid and well presented. It is not surprising to see a diverse outcome of risk documented in each case and reflects the nature of that institution, probably age of the facility, experience of staff and regulators and the strength of the regulating agency.

We agree

8-         The author’s, from their review of the 70 studies from 2020-2025 provided an analysis of risk assessment methodologies for PET radiopharmaceutical production focusing on occupational, technological and environmental risks. As they have highlighted, lot of this evidence was based on ‘probabilistic modeling, machine learning, multi-criteria decision-making, and failure mode and effects analysis’. This is a suitable starting point and source of information. However, an even greater source if such information is dose records- which the author’s clearly mentioned and very importantly information that can be gained from the regulatory/ licensing agencies that review, failures, incidents, non-conformances and other adverse events.

We agree

 

This review is a good starting point for future studies. The question many readers will ask – is ‘where do I fit in’ and what have I learnt from this and how do I implement best practice’’ . The question is ‘what is best practice’ and which of these 70 reviews describes or operates under ‘best practice’.

Your input has improved the manuscript's scientific quality and applicability.

Best regards,
Frank Montero Díaz (Corresponding Author, on behalf of the authors)

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The format used to cite the references in Table 1 needs to be revisited.

Author Response

Reviewer 2, Second round

1-The format used to cite the references in Table 1 needs to be revisited.
We sincerely appreciate your co"A Systematic Review of Integrated Risk Indicators for PET Radiopharmaceutical Production: Methodologies and Applications."nstructive feedback on our manuscript,  In response to your comment regarding the citation format in Table 1, we recognize the importance of consistency. This issue was revised to adopt a uniform author-year citation style for all 70 entries. This revision will ensure coherence with the manuscript’s reference list and improve overall readability. We believe these modifications have significantly enhanced the presentation of our data.