![]()
Design of a virtual simulator for micro-tasks training in microsurgery
Sector
Serious Games, Research, Surgery
Challenge
Creation of a virtual simulator to train micro surgeons
My role
Plan the process, research the user, define requirements, design solutions and test the design
Project time
6 months
Introduction
Using the Human Centered Design Process, over the course of 6 months during the summer 2020 we created a virtual reality simulator for microsurgery training.
About Microsurgery: Microsurgery is a surgical discipline that combines magnification with advanced diploscopes, specialized precision tools and various operating techniques. These techniques are primarily used to anastomose small blood vessels (arteries and veins) and to coapt nerves. Two of the main purposes of microsurgery are to transplant tissue from one part of the body to another and to reattach amputated parts.
This project was carried out in collaboration with the IBISC laboratory of the University Evry, and the Hospital: Centre Hospitalies Sud Francilien.
Understand and specify the user
We were also lucky to visit an institute in which microsurgery is taught. It gave us the opportunity to learn about the curriculum, ask questions to instructors and to get some hands on experience ourselves.
Microsurgery is mainly taught on rats since this is the "gold" standard of microsurgery education currently.
This has the advantage that students can learn on a living system, which behaves real, but brings in downsides as variable costs, no room to try out things, and ethical issues.
For an understanding of the user and the context of use, we observed a microsurgical intervention in the operating theatre.
Getting to know the environment in which the technique is applied helped in better empathise the needs of the users, but also facilitates future interviews and discussions with micorsurgeons.
Based on observations, interviews and a profound literature review, we created a questionnaire for teachers and another for students.
Aim of the questionnaire was to evaluate assumptions about their profiles. In the picture you can see the answers collected in a boxplot.
Two user profiles were created. One for students, and one for teachers.
Based on those profiles we could now conceptualise a system tailored to the needs of our users.
Of course it is not the goal to meet all the needs of the users, but rather solve some significant points well.
Finally we decided that a virtual reality system has the potential of meeting the needs best.
As an additional outcome of interviews and research conducted, following Cognitive Task Analysis was created. It decomposes the process of an anastomoses into its subtasks, including different decisions that have to be taken, and which instruments to use at each step. Several iterations of evaluations were done with the surgeons to validate the Cognitive Task Analysis presented.
Specify User Requirements
3 training modules were created to be prototyped and tested.
The module to the right is derived from a real excercise of microsurgery called "round the clock". It is an excercise that has a high degree of validation and which is well conceived for training. The idea is to prototype the task in virtual reality to see how the learned skills translate to the real excercise.
Following module called "microgrids" is derived from a warm up excercise for microsurgery students. Studies have shown that giving different degrees of difficulty for simulators helps in the development of the student.
The time limit of 20 seconds is taken from a paper published about the excercise. This permits to have comparable values once the testing starts.
We wanted to ellaborate further upon the excercise of microgrids and modify the squares to stars with a larger time window for the students to train the excercise.
Being able to follow straight lines seemed important to us since when dissecting under the microscope, it is also required to move the hand straight as precisely as possible. By following the outline of little stars we wanted to simulate such a situation.
Design and Development
How to succesfully create a virtual microscope was one of the first questions our mind was gravitating around from the start. Finally the solution was as easy as fixing a Head Mounted Display on a stand in a 45 degree angle. The user would bend over and look into the headset, without putting it on. For a better understanding check out the video.
To control the virtual tools, two haptic arms were put in place which deliver haptic feedback to the user and let's control the tools accurately, like you would be holding a real tool.
A microscope would be nothing without it's zoom. For the ease of prototyping we used the keys on the keyboard to zoom in, out and move around the environment if necessary.
The video on the left shows the "Round the Clock" task in action, and the right video explains the "Microgrids" task.
Testing
9 people tested all three prototypes. 1 expert, 2 intermediate and 6 novices who never worked under a microscope. Besides gathering feedback about the system, we were interessted to see if we can distinguish an expert from a novice in our environment.
It could be observed that the expert finished the task in 143 seconds, which was twice as fast as the average novice (283.5s) and intermediate (287.5s).
Plotting the movements of an expert versus a novice doing the Migrogrids task, looking from the top it can be seen how much more accurate the movements of the experts in filling the square were. Straight lines are visible, whereas the novice had a hard time drawing inside the square.
It is necessary to lift the tool to go to the next tile. Participants that lifted their tool more than 2.5 mm to change tile, performed in fact significantly (p=0.032) worse in the round the clock excercise than their counter peers.
The User Experience Questionnaire did not reveal any major problems of the prototype. The category dependability scored worst, which could be improved in the future by working out the ergonomics of the system and solving some minor bugs.
Conclusion
With our project we could show that it will be possible to build a microsurgery training simulator using a Head Mounted Display and two haptic arms. First doubts about the feasability due to technical issues as for example the accuracy of the arms could be eliminated. Although there are still things left open to be improved, this work can be taken as a basis for future iterations to deliver a novel high fidelity simulator to microsurgery students, which could enhance their future learning experience.