Troubleshooting in a Digital World—Server Failure of OIS in Radiotherapy from a Medical Perspective
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript presents a case study on managing patient treatment and workflow following the failure of an Oncology Information System (OIS). The Authors outline the steps taken to handle the situation and provide a detailed breakdown of the actions taken, highlighting the procedures and precautions necessary for such events.
I commend the Authors for choosing this topic, as a case study of this nature could prove extremely beneficial for other medical centers should a similar system failure occur. The step-by-step approach is well-documented, offering a clear pathway for responding to such incidents.
I recommend a Minor revision of the manuscript, primarily for stylistic choices and the absence of essential references.
My suggestions include:
1. Abbreviation Consistency and written quality: Ensure the consistent use of abbreviations such as OIS, HIS, and PACS. Repetition of terms in the phrases should also be minimized. Phrase length is good and overall the manuscript is very readable.
2. References:
- Number References: Instead of writing out the full names of sources, numbering them would make the manuscript easier to read and consult.
- General Definitions: Include references that explain what each system (OIS, HIS, PACS) is, to justify the definitions provided by the Authors.
For references, I recommend the following:
1. Collen, M.F. HIS concepts, goals and objectives. In Bakker, A.R. et al. (eds.) Towards New Hospital Information Systems.*Amsterdam: North Holland; 1988. pp. 3-9.
2. Bakker, A.R., Kouwenberg, J.M.L., Ottes, F.P. HIS and PACS integration aspects. In: MEDINFO 89 Part I, North Holland, Amsterdam: 1989; pp. 377-381.
3. Bakker, A.R. The Future of HIS; conclusions and recommendations of an IMIA Working Conference. PROCS 8th Annual Joint National Congress, The Practice of Medical Informatics, Hammond, Ed.W., Washington Ph.D., May 1989, American Association for Medical Systems and Informatics (AAMSI), San Francisco 1989, Vol. 7 pp. 230-234.
4. Lodwick, G.S., et al. Pictorial information systems and radiology improving the quality of communications. In Hoehne, K.K., (ed.) Pictorial Information Systems in Medicine. Berlin: Springer Verlag; 1984: pp. I-29.
5. Leguit, F.A. The Case: Leiden University Information System. In: PROCS of the IFIP IMIA Working Conference, Towards New Hospital Information Systems, Nijmegen, The Netherlands, Bakker, A.R., Ball, M.J., Scherrer, J.R., Willems, J.L. (eds.), Elsevier Science Publishers B.V. (North Holland), Amsterdam, pp. 321-326, ISBN: 0-444-70502-3.
6. Lodder, Herman, et al. "HIS-PACS coupling in practice." Medical Imaging III: PACS System Design and Evaluation. Vol. 1093. SPIE, 1989.
7. Bick, U., and H. Lenzen. "PACS: the silent revolution." European Radiology 9 (1999): 1152-1160.
In summary, the case study provides a guide for managing an OIS failure. Some revisions are advised to improve clarity and citation consistency. Nonetheless, I believe the manuscript could become a useful resource for medical professionals dealing with similar challenges.
Comments on the Quality of English LanguagePlease see comments to the Authors.
Author Response
- Abbreviation Consistency and written quality: Ensure the consistent use of abbreviations such as OIS, HIS, and PACS. Repetition of terms in the phrases should also be minimized. Phrase length is good and overall the manuscript is very readable. Replay: we tried to minimise it but it is important for the text (Corrected)
2. References:
- Number References: Instead of writing out the full names of sources, numbering them would make the manuscript easier to read and consult. (done)
- General Definitions: Include references that explain what each system (OIS, HIS, PACS) is, to justify the definitions provided by the Authors. (in the abbraviations). ((Corrected))
Reviewer 2 Report
Comments and Suggestions for AuthorsThe title “Trouble shooting in a digital world – Server Failure of OIS in Radiotherapy from a Medical Perspective” the title is fairly appropriate to the article topic.
This paper discusses a case study of an ARIA OIS server failure caused by an interface failure in the central IT department and the systematic approach developed to manage such failures from a clinical perspective. The steps taken to address the failure are described, including the immediate response and the long-term solutions implemented to prevent future occurrences. This manuscript, although providing a comprehensive guideline that other radiotherapy departments can adapt to improve their preparedness for similar events, lacks an introduction to the state-of-the-art imaging and tool usage that are now the basis of every step of the workflow from preclinical research to the clinic for both diagnosis and therapy.
Overall, the paper is acceptable, but it should be improved, including more state of the art evidence in the use of tools and software in practical activity from the bench to clinical bed.
More detailed comments are below.
Given the limited state of the art in this manuscript, it is appropriate to increase the range of the target readers, as the problem discussed in the article must also be known to those working in the field of developing new tools or using artificial intelligence software and algorithms in the biomedical field. In [10.1007/978-3-031-51026-7_9] the authors applied a standardized and reproducible computational statistical analysis model and identified predictive and prognostic models to facilitate the process of medical decision-making based on one of the most widely used software in the biomedical field, namely PyRadiomics, whose use is widely spread in radiology and radiotherapy for the staging of tumors, such as bladder cancer. By distinguishing low-grade (LG) from high-grade (HG) bladder lesions and non-muscle invasive bladder cancer (NMIBC) from muscle invasive bladder cancer, this work has improved the characterization and differentiation of bladder lesions, both in terms of differentiation of LG lesions from HG and discrimination between NMIBC and MIBC. This has enabled medical doctors and radiation oncologists to subsequently plan specific treatment plans that would increase performance and specificity, not only in cancer staging, but above all in future treatment plans. This article should be strongly considered in discussions and future perspectives, as often in the research world the possible obstacles associated with the use of software and computer interfaces are not considered enough. This could help establish clear communication protocols, validate reliable backup systems and promote cooperation with radiotherapy departments and research teams operating in the field in order to minimize interruptions of patient care during software failures at any level.
In [10.1038/s41467-022-30728-3] a significant interview with experts was conducted. From this manuscript it is clear that the challenges that concern Problem solving in a digital world from a medical perspective include various factors that could be addressed a priori, such as poor or inadequate infrastructure, slow or intermittent Internet connectivity, inequalities in electricity availability. If processes are not mapped and improved or reviewed regularly, digitalization becomes a dispersive and chaotic process that inevitably leads to system errors and increase problems. As a result, this article should be considered to better explain and defend such a complex and impactful document, with a view to offering other hypothetical solutions to prevent the onset of operating system errors within hospitals.
The figures 1, 2 and 3 appear to be of low resolution. Please replace them with high resolution figures with more attractive colors for viewing.
To make reading the document easier for the reader, an Abbreviations section should be added to the end of the document.
In conclusion, to make the article more coherent, the authors should expand the references and improve the English to make the discourse more fluent throughout the document.
Comments on the Quality of English Languagethe authors should improve the English to make the discourse more fluent throughout the document.
Author Response
Given the limited state of the art in this manuscript, it is appropriate to increase the range of the target readers, as the problem discussed in the article must also be known to those working in the field of developing new tools or using artificial intelligence software and algorithms in the biomedical field. In [10.1007/978-3-031-51026-7_9] the authors applied a standardized and reproducible computational statistical analysis model and identified predictive and prognostic models to facilitate the process of medical decision-making based on one of the most widely used software in the biomedical field, namely PyRadiomics, whose use is widely spread in radiology and radiotherapy for the staging of tumors, such as bladder cancer. By distinguishing low-grade (LG) from high-grade (HG) bladder lesions and non-muscle invasive bladder cancer (NMIBC) from muscle invasive bladder cancer, this work has improved the characterization and differentiation of bladder lesions, both in terms of differentiation of LG lesions from HG and discrimination between NMIBC and MIBC. This has enabled medical doctors and radiation oncologists to subsequently plan specific treatment plans that would increase performance and specificity, not only in cancer staging, but above all in future treatment plans. This article should be strongly considered in discussions and future perspectives, as often in the research world the possible obstacles associated with the use of software and computer interfaces are not considered enough. This could help establish clear communication protocols, validate reliable backup systems and promote cooperation with radiotherapy departments and research teams operating in the field in order to minimize interruptions of patient care during software failures at any level.
In [10.1038/s41467-022-30728-3] a significant interview with experts was conducted. From this manuscript it is clear that the challenges that concern Problem solving in a digital world from a medical perspective include various factors that could be addressed a priori, such as poor or inadequate infrastructure, slow or intermittent Internet connectivity, inequalities in electricity availability. If processes are not mapped and improved or reviewed regularly, digitalization becomes a dispersive and chaotic process that inevitably leads to system errors and increase problems. As a result, this article should be considered to better explain and defend such a complex and impactful document, with a view to offering other hypothetical solutions to prevent the onset of operating system errors within hospitals.
The figures 1, 2 and 3 appear to be of low resolution. Please replace them with high resolution figures with more attractive colors for viewing. this is best resolution , colors to change is difficult
To make reading the document easier for the reader, an Abbreviations section should be added to the end of the document. in the end we have it in the appendix
In conclusion, to make the article more coherent, the authors should expand the references and improve the English to make the discourse more fluent throughout the document. done (Abbreviations section is added and improved )
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors do not answer the first two questions, they answer the next three but by inserting the answers inside the text of the comment and do not highlight in yellow the changes they have made to the text.
The authors must answer using this format:
comment 1.
answer 1.
They must indicate in the answers the lines where the changes have been made and highlight in yellow the changes in the text.
They must also answer the first two comments integrating the suggested references.
They must highlight in yellow all the changes in the text.
For clarity I report the two comments to which they have not answered and integrated the references:
Given the limited state of the art in this manuscript, it is appropriate to increase the range of the target readers, as the problem discussed in the article must also be known to those working in the field of developing new tools or using artificial intelligence software and algorithms in the biomedical field. In [10.1007/978-3-031-51026-7_9] the authors applied a standardized and reproducible computational statistical analysis model and identified predictive and prognostic models to facilitate the process of medical decision-making based on one of the most widely used software in the biomedical field, namely PyRadiomics, whose use is widely spread in radiology and radiotherapy for the staging of tumors, such as bladder cancer. By distinguishing low-grade (LG) from high-grade (HG) bladder lesions and non-muscle invasive bladder cancer (NMIBC) from muscle invasive bladder cancer, this work has improved the characterization and differentiation of bladder lesions, both in terms of differentiation of LG lesions from HG and discrimination between NMIBC and MIBC. This has enabled medical doctors and radiation oncologists to subsequently plan specific treatment plans that would increase performance and specificity, not only in cancer staging, but above all in future treatment plans. This article should be strongly considered in discussions and future perspectives, as often in the research world the possible obstacles associated with the use of software and computer interfaces are not considered enough. This could help establish clear communication protocols, validate reliable backup systems and promote cooperation with radiotherapy departments and research teams operating in the field in order to minimize interruptions of patient care during software failures at any level.
In [10.1038/s41467-022-30728-3] a significant interview with experts was conducted. From this manuscript it is clear that the challenges that concern Problem solving in a digital world from a medical perspective include various factors that could be addressed a priori, such as poor or inadequate infrastructure, slow or intermittent Internet connectivity, inequalities in electricity availability. If processes are not mapped and improved or reviewed regularly, digitalization becomes a dispersive and chaotic process that inevitably leads to system errors and increase problems. As a result, this article should be considered to better explain and defend such a complex and impactful document, with a view to offering other hypothetical solutions to prevent the onset of operating system errors within hospitals.
Minor editing of English language required.
Author Response
We thank the reviewers for their insightful and constructive comments, which have greatly helped us to improve our manuscript. Below are our detailed responses.
Reviewer Comment 1
Given the limited state of the art in this manuscript, it is appropriate to increase the range of the target readers, as the problem discussed in the article must also be known to those working in the field of developing new tools or using artificial intelligence software and algorithms in the biomedical field. In [10.1007/978-3-031-51026-7_9] the authors applied a standardized and reproducible computational statistical analysis model... This article should be strongly considered in discussions and future perspectives..."*
Response:
We greatly appreciate this recommendation and have expanded the scope of the discussion section to emphasize the interdisciplinary relevance of our work. As suggested, we have now included a citation and brief discussion of the study referenced by the reviewer (DOI: [10.1007/978-3-031-51026-7_9]). This work highlights the successful implementation of reproducible radiomics pipelines in oncology using PyRadiomics and underscores how robust digital infrastructures—especially OIS and TPS platforms—are essential for the reproducibility and transferability of AI-driven solutions.
In the **Discussion and Conclusion** sections, we now highlight that software outages not only impact clinical workflows but also interrupt AI-based systems that depend on consistent access to high-quality, structured medical data. This connection emphasizes the broader implications of OIS failures and underscores the need for system-level resilience, particularly when integrating radiomics and machine learning approaches into clinical practice.
Additionally, we elaborate on the potential of standardized protocols and improved data pathways to facilitate cooperation between radiation oncology departments and research teams focused on tool development. These additions aim to broaden the readership to include experts in biomedical informatics and AI, as advised.
Reviewer Comment 2:
In [10.1038/s41467-022-30728-3] a significant interview with experts was conducted... From this manuscript it is clear that the challenges that concern Problem solving in a digital world from a medical perspective include various factors... This article should be considered to better explain and defend such a complex and impactful document...
Response:
Thank you for this valuable suggestion. We have now incorporated a reference to the article mentioned by the reviewer (DOI: [10.1038/s41467-022-30728-3]) and discussed its relevance in the **Introduction** and **Discussion** sections. Specifically, we acknowledge that the success of digital infrastructures in radiotherapy is inherently tied to broader systemic issues, such as infrastructure quality, internet connectivity, and hardware/software resilience.
In response, we now explicitly mention that sustainable digitalization in oncology cannot rely solely on clinical protocols or emergency plans. Instead, it requires the simultaneous development of institutional strategies for digital governance, regular mapping of digital processes, and integration of disaster resilience planning. These elements, discussed in the cited work, further support the rationale for our proposed escalation levels and the emphasis on proactive testing and interdepartmental coordination.
Moreover, we now highlight in our Conclusion that system errors often stem from "unseen" infrastructural weaknesses. Addressing these weaknesses requires both medical and administrative cooperation, as well as awareness and planning that extends beyond individual departments. This addition helps position our manuscript within a broader context of digital health resilience.
We are grateful for the reviewers' input, which helped us substantially strengthen our manuscript. We hope that the revisions meet the expectations and are happy to make further clarifications if required.
Reviewer Comment 3:
The figures 1, 2 and 3 appear to be of low resolution. Please replace them with high resolution figures with more attractive colors for viewing.
response:
this is best resolution , colors to change is difficult
Reviewer comment 4:
To make reading the document easier for the reader, an Abbreviations section should be added to the end of the document. in the end we have it in the appendix
In conclusion, to make the article more coherent, the authors should expand the references and improve the English to make the discourse more fluent throughout the document.
response:
done (Abbreviations section is added and improved )
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Comments and Suggestions for Authorsthe authors say to take into consideration the suggested topic fully discussed by the authors of this original paper applied on a dataset of patients, mentioning the DOI [10.1007/978-3-031-51026-7_9] and then insert a generic reference to a review that although dealing with a similar but not identical topic does not show the workflow necessary for the application of RADIOMICS in the clinical setting.
it is requested to insert this reference [10.1007/978-3-031-51026-7_9] inside the introduction.
Comments on the Quality of English LanguageMinor editing of English language required.
Author Response
We thank the reviewer for their insightful comment and for highlighting the importance of including the relevant reference. In response, we have incorporated a direct citation to the work (DOI: [10.1007/978-3-031-51026-7_9]) in the Introduction section of the manuscript. This reference illustrates a comprehensive radiomics workflow and complements our discussion by emphasizing the clinical relevance and technical requirements of digital data infrastructures in radiotherapy. We agree that including this source improves the clarity and completeness of our argument regarding the significance of robust digital systems in enabling advanced applications like radiomics.