Needle Transportable Semi-Automatic Hair Follicle Implanter and Image-Based Hair Density Estimation for Advanced Hair Transplantation Surgery

Round 1

Reviewer 1 Report

The language used looks like some of it was taken from a patent application that the applicant might have applied for as it is technically sophisticated but not easy to follow. It's style is a legalistic/engineering style, which will be difficult for non-engineers to follow. As a surgeon who has many patents, I understood the logic in the language used; however, most physicians might skip over it because of its technical nuances. 

Author Response

Ans)

1) First of all, the title has been changed to match the content as shown below.

(Previous) Needle Transportable Automatic Hair-follicle Implanter and Image Based Hair Density Estimation for Robotic Hair Transplantation

(Current) Needle Transportable Semi-Automatic Hair Follicle Implanter and Image Based Hair Density Estimation for Advanced Hair Transplantation Surgery

In Figure 6 and Figure 10, Several terms have been changed more easily. (For example, support part -> needle holder, needle aligner -> needle mover)

2) And, to explain the developed system in more detail, the flowchart (Figure 7) and the developed motor control program (Figure 12) have been added as the following.

Figure 7. Operation flowchart for continuous needle supply and follicle transplantation.

(line 166) Figure 7 shows the operation flowchart for continuous needle supply and HF transplantation in the proposed semi-automatic implanter. When the NSM equipped with multiple needles is combined with the semi-automatic implanter body, the needles are sequentially supplied to the needle gripper. A single HF transplant is performed through the forward and backward movement of the needle gripper. After all the needles in NSM are used, a new NSM is supplied.

Figure 12. Developed motor control program using the EPOS2 digital positioning controller.

(line 246) The motor control program currently supports three motors, but it only uses two for the semi-automatic implanters. It has a selectable gear ratio and can set the rotation speed, acceleration and/or deceleration speed, and delay time. Since the motor encoder signal is 1024/cycle, the speed setting supports a resolution of 1/1024. The motor control program also has a homing mode function when both motors stop working. Its command comprise rotation direction, number of rotations, speed of rotation, acceleration, deceleration, delay time, and gear ratio. In addition, it can store an entirely new command or call up an existing command. Each motor is configured with a USB port interface with a baud rate and timeout of 106 and 500 ms, respectively.

3) We improved the English and style of the paper through native speaker.

Author Response File: Author Response.docx

Reviewer 2 Report

This is a very interesting interdisciplinary paper about the automatic hair-follicle implantation, but there are some questions which require clarification.

The title indicates “Robotic Hair Transplantation”, but in this case only a semi-automatic hair-follicle implanter was used (without the use of robot).

There is only a short mention that “robotic system was developed to harvest automatically FUs”.

I guess that this semi-automatic hair implanter can be connected to a surgery robot in order to reduce the fatigue of the surgeon, therefore the title should be changed and some comments about the further work can be added in conclusions.

 

Is that true -  “The total distance travelled by a surgeon's arm, moving the manual planter per patient, is close to 1 km”?

The methodology used is described in detail, but the statistical analysis of results raises some doubts and should be improved.

There are only 6 cases, which are too few for a detailed statistical analysis.

For a paired T-test 10 cases are recommended. Some specialists allow fewer samples, but for this test to be valid the differences need to be normally distributed [38]. Also, hypothesis H0,  H1 and the calculation of the t-value should be described.

Pearson’s correlation can only be used for independent variables (for example blood pressure and hair density), but not in this case where variables are depended (therefore the paired T-test for depended variables is commonly used in similar cases). 

What exactly does “1.33 high engraftment rates in the automatic implanter ” represent?

 

In order to represent the results clearly the percentage can be used as in point 4.1.

If we have 25 implanted hair- follicles in each case, then an average rate of 60,8% for automatic implanter and 55,2% for manual implanter is obtained. There is a difference of 5,6% engraftment rate for the automatic implanter.

(The rate is also much lower than in the mice experiment.)

The question is – Is this statistically significant? That should be proved.

 

Improvement of the English language and style is required.

Author Response

This is a very interesting interdisciplinary paper about the automatic hair-follicle implantation, but there are some questions which require clarification. The title indicates “Robotic Hair Transplantation”, but in this case only a semi-automatic hair-follicle implanter was used (without the use of robot). There is only a short mention that “robotic system was developed to harvest automatically FUs”. I guess that this semi-automatic hair implanter can be connected to a surgery robot in order to reduce the fatigue of the surgeon, therefore the title should be changed and some comments about the further work can be added in conclusions.

Ans) Based on your advice, the title has been changed as follows. And what you mentioned was added to the conclusion.

(Previous) Needle Transportable Automatic Hair-follicle Implanter and Image Based Hair Density Estimation for Robotic Hair Transplantation

(Current) Needle Transportable Semi-Automatic Hair Follicle Implanter and Image Based Hair Density Estimation for Advanced Hair Transplantation Surgery

(Conclusion, line 477) Furthermore, it can be connected to a surgical robot and utilized in a fully automatic hair transplantation system.

 

Is that true - “The total distance travelled by a surgeon's arm, moving the manual planter per patient, is close to 1 km”?

Ans) It was inferred by assuming that distance from the manual follicular site loaded with the hair follicle by the nurse to the patient's scalp is 50 cm, and the number of hair follicle transplants in one patient procedure is 2000 follicles. I think this analysis is misleading. So, I modified the sentence as follows.

(437) Owing to thousands of repetitive movements per patient, surgeons feel much fatigue in surgery, which can affect the overall surgical outcome.

 

The methodology used is described in detail, but the statistical analysis of results raises some doubts and should be improved. There are only 6 cases, which are too few for a detailed statistical analysis. For a paired T-test 10 cases are recommended. Some specialists allow fewer samples, but for this test to be valid, the differences need to be normally distributed [38]. Also, hypothesis H0, H1 and the calculation of the t-value should be described.

Ans) Your explanation is correct. The following description has been added.

(line 377) Originally, there were 8 clinical subjects, but according to the clinical trial plan, a total of 6 patients completed the clinical trial, except for 1 dropout due to hair transplant failure (device malfunction during surgery) and 1 dropout due to IRB recommendation. The null hypothesis (H0) is that the clinical effectiveness of the semi-automatic implanter will be different from that of the manual implanter. And the alternative hypothesis (H1) is that the clinical effectiveness of the semi-automatic implanter will not differ from that of the manual implanter. The difference between the HFs regrown by the manual implanter and the semi-automatic implanter satisfied the normality in both Kolmogorov-Smirnov and Shapiro-Wilk as the significance probability was higher than the significance level (p=0.06) as shown in Table 4.

 

Pearson’s correlation can only be used for independent variables (for example blood pressure and hair density), but not in this case where variables are depended (therefore the paired T-test for depended variables is commonly used in similar cases).

Ans) Regarding the independent variable relationship to Pearson correlation, you are right. However, (line 397) this study assumed that the mechanically operated semi-automatic implanter and the manual implanter operated by human experience are independent. Therefore, the Pearson correlation analysis of this study may not be accurate. We have added this to the paper.

 

What exactly does “1.33 high engraftment rates in the automatic implanter ” represent? “ In order to represent the results clearly the percentage can be used as in point 4.1.

Ans) That represents the average number of regrown hair follicles. Following your advice, the sentences were revised as follows. Also, the percentage of engraftment rate was added in Table 3.

(line 385) The result of the statistical analysis shows that the average number of regrown HFs by the semi-automatic implanter was 1.33 (6% higher engraftment rate), more than that of the manual implanter (p = 0.062).

(line 403) However, this is due to an average 6% higher engraftment rate in the semi-automatic implanter.

 

If we have 25 implanted hair- follicles in each case, then an average rate of 60,8% for automatic implanter and 55,2% for manual implanter is obtained. There is a difference of 5,6% engraftment rate for the automatic implanter. (The rate is also much lower than in the mice experiment.) The question is – Is this statistically significant? That should be proved.

Ans) the following explanation is added.

(line 413) To date, the HF engraftment rate by manual implant surgery is considered very high [21]. In our experiments, the semi-automatic implanter showed 6% higher HF engraftment rates than the manual implanter. This figure may not be statistically meaningful, but it is important in that it confirms the possibility that delicate procedures of existing manual implanters can be automated. The actual engraftment rate in clinical experiments is lower than that in mice experiments because rats lack immunity, whereas the human scalp is significantly affected by individual immunity as well as the condition of HFs and scalp.

 

Improvement of the English language and style is required.

Ans) We improved the English and style of the paper.

Author Response File: Author Response.docx

Reviewer 3 Report

The topic of the paper is interesting. However, the paper has to be significantly improved. The main idea would be accetable, but paper has to be rewritten in several aspects. 

There are some my suggestions to the authors:

1. You should not use abbrevitations in the abstract. Use whole titles/names and use abbrevitations after the abstract.
2. Figure 2: you write about disadvantages of hair transplantation using manual implanter. Is it measurable how these disadvantages decrease using your approach ? IS it possible to describe it by any parameters ?
3. What about previous solutions in this area ? Besides classical manual implanters. You should describe also other solutions in order to reader be able to compare them with your solution.
4. Figure 4: Low quality of the figure. Should be changed. 
5. Maybe would be helpful for reader to see any flowchart of automatic implanter, how it works. Show basic principles.
6. Page 5: you describe automatic implanter. I am sorry if I am wrong, but from text it is not clear if automatic implanter from figure 5 is your system or not. Should be clearly point out. 
7. If it is your system, describe it in more details. For example describe control system, electronic system, etc. 
8. Figure 6-7: Low quality of the figure. Should be changed.
9. You write that automatic implanter consists of maxon motors. Describe some control algorithm by which these motors work. 
10. Figure 8: low quality of the figure.
11. Figure 9: describe in more details electronic and control system shown in this figure.
12. Figure 12: there are not labels for the axes (units). But quality of the figure is very poor again.
13. The conclusion is written in not very suitable way. You have to emphasize your own contributions. From the introduction and background chapters, the readers should have any basic knowledge from this area with focus on previous solutions. But from your introduction and background it is not clear. It has to be rewritten. Then your conclusion should consists of the facts, from which it should be clear if your work is beneficial, if you did some contribution.
14. The concept of whole paper should be modified, because it is a little be confusing now. 
15. You introduced some device. There so many things which could be described, from many views. But your description is a little bit poor, without any mathematical / control / background.

 

Author Response

The topic of the paper is interesting. However, the paper has to be significantly improved. The main idea would be acceptable, but paper has to be rewritten in several aspects.

There are some my suggestions to the authors:

1. You should not use abbrevitations in the abstract. Use whole titles/names and use abbrevitations after the abstract.

Ans) The abbrevitations (HF, GMP) were removed from the abstract.

 

2. Figure 2: you write about disadvantages of hair transplantation using manual implanter. Is it measurable how these disadvantages decrease using your approach ? IS it possible to describe it by any parameters ?

Ans) Based on your point of view, the explanation below has been added under Figure 3.

(line 105) In the current hair transplant surgery, a minimum of 2000 HFs are implanted in one patient. In each transplant, the surgeon executes a series of movements such as gripping, moving, aiming, transplanting, and in situ movement of a manual implanter. Repetition of these actions directly causes fatigue.

 

3. What about previous solutions in this area ? Besides classical manual implanters. You should describe also other solutions in order to reader be able to compare them with your solution.

Ans) We added the following explanation.

(line 108) Although several advanced devices for continuously extracting HFs, such as P-FUE and ARTAS, have been invented, devices for continuously transplanting HFs have not been developed yet. This is because the operation of the HF transplant is much more sophisticated and complicated than the HF extraction.

 

4. Figure 4: Low quality of the figure. Should be changed.

Ans) The figure has been modified with high resolution.

 

5. Maybe would be helpful for reader to see any flowchart of automatic implanter, how it works. Show basic principles.

Ans) The following flowchart and its description have been added.

Figure 7. Operation flowchart for continuous needle supply and follicle transplantation.

(line 166) Figure 7 shows the operation flowchart for continuous needle supply and HF transplantation in the proposed semi-automatic implanter. When the NSM equipped with multiple needles is combined with the semi-automatic implanter body, the needles are sequentially supplied to the needle gripper. A single HF transplant is performed through the forward and backward movement of the needle gripper. After all the needles in NSM are used, a new NSM is supplied.

 

6. Page 5: you describe automatic implanter. I am sorry if I am wrong, but from text it is not clear if automatic implanter from figure 5 is your system or not. Should be clearly point out.

Ans) Yes it is. Automatic implanter was developed by us. The following description has been added.

(line 151) The structure of the semi-automatic implanter developed in this study is shown in Figure 6 [26].

 

7. If it is your system, describe it in more details. For example describe control system, electronic system, etc.

Ans) For the proposed system, the flowchart (Figure 7) and the developed motor control program (Figure 12) have been added.

 

8. Figure 6-7: Low quality of the figure. Should be changed.

Ans) The figure has been modified with high resolution.

9. You write that automatic implanter consists of maxon motors. Describe some control algorithm by which these motors work.

Sol) The following content and figure have been added.

Figure 12. Developed motor control program using the EPOS2 digital positioning controller.

(line 246) The motor control program currently supports three motors, but it only uses two for the semi-automatic implanters. It has a selectable gear ratio and can set the rotation speed, acceleration and/or deceleration speed, and delay time. Since the motor encoder signal is 1024/cycle, the speed setting supports a resolution of 1/1024. The motor control program also has a homing mode function when both motors stop working. Its command comprise rotation direction, number of rotations, speed of rotation, acceleration, deceleration, delay time, and gear ratio. In addition, it can store an entirely new command or call up an existing command. Each motor is configured with a USB port interface with a baud rate and timeout of 106 and 500 ms, respectively.

 

10. Figure 8: low quality of the figure.

Ans) The figure has been modified with high resolution.

 

11. Figure 9: describe in more details electronic and control system shown in this figure.

Sol) For Figure 11 (previously Figure 9), the flowchart (Figure 7) and the developed motor control program (Figure 12) have been added.

 

12. Figure 12: there are not labels for the axes (units). But quality of the figure is very poor again.

Sol) The axis label and quality of the picture have been modified as below.

(a)

(b)

Figure 14. (a) Consistency and (b) comparison of engraftment rate.

 

13. The conclusion is written in not very suitable way. You have to emphasize your own contributions. From the introduction and background chapters, the readers should have any basic knowledge from this area with focus on previous solutions. But from your introduction and background it is not clear. It has to be rewritten. Then your conclusion should consists of the facts, from which it should be clear if your work is beneficial, if you did some contribution.

Sol) Based on the reviewer's advice, the conclusion was revised as follows.

(line 466) The current hair transplantation procedure is divided into two stages: follicle extraction and follicle transplantation. The follicle extraction technique is automated using a motor and image processing techniques, such as P-FUE and ARTAS. Meanwhile, the follicle transplant technique mostly uses simple tools, such as FUE tweezers and FUT’s manual implanter, with relatively less automation. Follicle extraction is easy to automate because it is completed by a simple punching operation, whereas follicle transplantation is not easy to automate because of various considerations, such as implant depth, speed, and direction. Therefore, this study proposes a semi-automatic implanter structure to improve the method of completely manual follicular transplantation. The semi-automatic implanter was developed based on the long experience and know-how of hair transplantation surgeons. The developed semi-automatic implanter system allows even novice surgeons to obtain a high engraftment rate, which is equivalent to that of highly skilled surgeons. It may also reduce the cost of surgery as the hair transplantation procedure is simplified. Furthermore, it can be connected to a surgical robot and utilized in a fully automatic hair transplantation system.

 

14. The concept of whole paper should be modified, because it is a little be confusing now.

Ans) The concept of the original paper was revised in Abstract, Introduction, and Conclusion.

 

15. You introduced some device. There so many things which could be described, from many views. But your description is a little bit poor, without any mathematical / control / background.

Ans) In addition to the motor control system, the answer to your ninth question, the following background has been added.

 

(a)

(b)

(c)

Figure 3. Survey on development of automated hair implanter; (a) hair transplantation method, (b) surgery time, and (c) Willingness to use automated equipment.

On May 10, 2014, at the Korean Society of Hair Transplantation, a questionnaire was conducted on an automated hair transplantation device for a hair transplant clinician as shown in Figure 3. The total number of responses was 94, and 62% and 41% of respondents had clinical experience in hair transplantation for 3 years and 5 years or more, respectively. As a result of the survey, the proportion of using manual implanter in the method of transplanting hair was 87.2%, the time for transplanting 3000 HFs was 4~5 hours, and the intention to use an automated hair transplant device was 84%. Through the survey results, many doctors demand the need for an automated hair transplantation device, and if it is commercialized, it is judged that it will be able to create a sufficient market.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The improved version of the paper is now more readable and suitable for publication.

Reviewer 3 Report

Thank you for processing the suggestions. The paper has been improved.