Artificial Intelligence and the Medical Physicist: Welcome to the Machine

Round 1

Reviewer 1 Report

I read the paper with interest. The work is a valuable addition to the literature on AI/Medicine and highlights the role of physicists in this area. I recommend publication of the work after the below revisions:

Specific comments:

• In the address list, please remove an extra “1”:

1Medical Physics Department Centro ….

• First paragraph of the Introduction is too short. Please connect with the subsequent paragraph.
• In the second paragraph, please change uppercase A to lower case:

few years, And the knowledge …

• Connected to the second paragraph, please also discuss emerging areas at the interface of physics and medicine. Traditionally, physics contributed to medicine through instrumentation (e.g., imaging technologies etc) and AI is helpful in this regard (e.g., in image processing). However, in the future, we envision that concepts from various areas of physics including statistical physics will be used to address medical diagnostic problem and AI will also play a key role in such developments. Please refer to the below article and references therein:

Statistical Physics for Medical Diagnostics: Learning, Inference, and Optimization Algorithms

Diagnostics 2020, 10(11), 972; https://doi.org/10.3390/diagnostics10110972

 

• Also in section 5.8., it would be helpful to discuss how statistical physicists can contribute to medical diagnostic problems in the long term (Again see the above paper that reviewed this topic).

 

• One topic that I miss in the imaging section is the new development in Mass Spectrometry Imaging, which will emerge as a useful tool for clinical decision making in the future. Pathological specimen will be imaged using mass spectrometry and their proteins and metabolites will be coloured/mapped. Such images are rich in (high dimensional) data and thousands of proteins and metabolites might be imaged in one slide. See below sources:

Spatial Metabolomics and Imaging Mass Spectrometry in the Age of Artificial Intelligence

https://www.annualreviews.org/doi/10.1146/annurev-biodatasci-011420-031537

Automated mass spectrometry imaging of over 2000 proteins from tissue sections at 100-μm spatial resolution

https://www.nature.com/articles/s41467-019-13858-z

 

 

 

Author Response

We would like to thank the editors and reviewers for their valuable comments. Below, we provide point-by-point responses to their comments. We hope that the revised manuscript is acceptable for publication.

Reviewer #1

I read the paper with interest. The work is a valuable addition to the literature on AI/Medicine and highlights the role of physicists in this area. I recommend publication of the work after the below revisions:

Specific comments:

Comment: In the address list, please remove an extra “1”: 1Medical Physics Department Centro ….

Response: Done

Comment: First paragraph of the Introduction is too short. Please connect with the subsequent paragraph.

Response: The first paragraphs of the Introduction have been revised in order to be more easy to follow.

 

Comment: In the second paragraph, please change uppercase A to lower case: few years, And the knowledge…

Response: done

Comment: Connected to the second paragraph, please also discuss emerging areas at the interface of physics and medicine. Traditionally, physics contributed to medicine through instrumentation (e.g., imaging technologies etc) and AI is helpful in this regard (e.g., in image processing). However, in the future, we envision that concepts from various areas of physics including statistical physics will be used to address medical diagnostic problem and AI will also play a key role in such developments. Please refer to the below article and references therein:

Statistical Physics for Medical Diagnostics: Learning, Inference, and Optimization Algorithms

Diagnostics 2020, 10(11), 972; https://doi.org/10.3390/diagnostics10110972

Also in section 5.8., it would be helpful to discuss how statistical physicists can contribute to medical diagnostic problems in the long term (Again see the above paper that reviewed this topic).

Response: a paragraph on statistical physics was added to the introduction, citing the reference above, which reads: “Moreover, analytical and computational techniques of physics, in particular from statistical physics of disordered systems can be on large-scale problems including machine learning, e.g. to analyze the weight space of deep neural networks[6] [7].”

Comment: One topic that I miss in the imaging section is the new development in Mass Spectrometry Imaging, which will emerge as a useful tool for clinical decision making in the future. Pathological specimen will be imaged using mass spectrometry and their proteins and metabolites will be coloured/mapped. Such images are rich in (high dimensional) data and thousands of proteins and metabolites might be imaged in one slide. See below sources:

Spatial Metabolomics and Imaging Mass Spectrometry in the Age of Artificial Intelligence

https://www.annualreviews.org/doi/10.1146/annurev-biodatasci-011420-031537

Automated mass spectrometry imaging of over 2000 proteins from tissue sections at 100-μm spatial resolution

https://www.nature.com/articles/s41467-019-13858-z

Response: we added a paragraph on spatial metabolomics and mass spectrometry, citing these references: “Spatial metabolomics is a new field aiming at measuring the distribution of molecules such as  metabolites, lipids, and drugs within body structures, using imaging such as mass spectrometry, where each pixel is represented by its mass spectrum[70]. Being characterized by large amount of highly dimensional data, including overlapping and noisy molecular signals, this technique, looks promising for application of AI [71]”

Reviewer 2 Report

Authors have demonstrated a review on  how AI, machine learning techniques are important in medical industry. Since AI has been widely studied in current literature, it is hard to find the main approach Ans of the authors. Most imporatantly, this paper clearly reveals the gathered vital information.However, the next successful approach or the proposalof the authors is unclear. Obviously,a review must be contained with existing methods. But the next scope has to be well defined. And also,I suggest authors to bring up few qualitative information to enhance th readability and the understandability of the paper.The   

Author Response

We would like to thank the editors and reviewers for their valuable comments. Below, we provide point-by-point responses to their comments. We hope that the revised manuscript is acceptable for publication.

Reviewer #2

Comment: Authors have demonstrated a review on  how AI, machine learning techniques are important in medical industry. Since AI has been widely studied in current literature, it is hard to find the main approach Ans of the authors. Most importantly, this paper clearly reveals the gathered vital information. However, the next successful approach or the proposal of the authors is unclear. Obviously, a review must be contained with existing methods. But the next scope has to be well defined. And also, I suggest authors to bring up few qualitative information to enhance the readability and the understandability of the paper. 

Response: We thank the reviewer for this comment and revised the manuscript by clarifying its purpose. In the abstract, the purpose is clarified as follows: “ The purpose of this review is to summarize the main applications of AI in Medical Physics, describe the skills of the MPs in research and clinical applications of AI, and define the major challenges of AI in healthcare”. Moreover, the manuscript was revised to increase its quality by avoiding repetitions, with the goal of emphasizing the relevant information.  

Reviewer 3 Report

A very good work that summarizes the value and potential of artificial intelligence in medicine. I also appreciate the limitations of AI presented and the tasks for scientists who will use AI.

Congratulations to all Authors

Author Response

We would like to thank the editors and reviewers for their valuable comments. Below, we provide point-by-point responses to their comments. We hope that the revised manuscript is acceptable for publication.

Reviewer #3

Comment: A very good work that summarizes the value and potential of artificial intelligence in medicine. I also appreciate the limitations of AI presented and the tasks for scientists who will use AI.

Response: We thank the reviewer for his/her kind words and hope that the revised manuscript is suitable for publication.

Round 2

Reviewer 1 Report

I am satisfied with the revisions. My comments have been addressed.