AI in Robotic Surgery
Hi guys and welcome back to the series on AI!
Today i will be focusing on how AI can transform surgery, which I hope you will find interesting.
How robotic surgery was used in the past:
Initially robotic surgery was used for carrying out surgical tasks that can be repetitive, the most common examples of this include suturing and tissue dissection. This increased consistency and also reduced surgeon workload so they could focus their attention on other parts of their job that requires more attention.
However, it is now being used for many other purposes.
Current use of AI in robotic surgery:
It has been proven to be particularly useful in image recognition
Identifying critical structures, for example blood vessels and tumours
Helping the surgeon with desicion making
What AI is hoping to be used for in surgery in the future:
Assist in planning and optimising surgical instrument movement
Smoother and more precise surgery
Give surgeon feedback on tissue texture and resistance
Personalise surgical responses to the patient's needs
There are 2 main AI algorithms being explored for surgery:
Deep Learning- image recognition, identification, prediction of bleeding risks
Reinforcment learning- this is trial and error method learning in the hope that the robot can perform complex tasks in the future
There are 5 different levels that AI can support a surgery in that will be explored below:
Level 0- surgeons directly controlling robotic movement without asistance
Level 1- assist surgeons by actively guiding them through limitations that are faced in certain surgeries, for example landoscopy through their active technology eg. tissue sensitivity
Level 2- execute specific surgical tasks based on physican guidelines
Level 3- plan and execute tasks independent of surgical setting examples include assesing thickness of tissue and planning for suture intersection. This is currently what AI is being used most within surgery
Level 4- interpreting pre-operative and intra-operative data to create intervention plans to execute actions adapting it to real time eg. targeting cancerous areas and avoiding healthy tissue
Level 5- surgery without human intervention. This is not done yet but, an aim for the future.
Advantages of AI in robotic surgery:
Enhanced precision and accuracy, specifically in delicate procedures
Reducing surgeon fatigue by carrying out repetitive tasks so surgeons can increase focus on critical aspects
Increased safety as their is a lower risk of bleeding, surgical errors, faster recovery time and less pain
Disadvantages of AI in surgery:
There is a very high intial cost which would require a large amount of the NHS budget. It is also less accessible to small hospitals and developing countries that do not have the resources for this technology
The reliability and beneficial outcome of AI is highly dependent on quantity and quality of its training
Cybersecurity concerns- if AI has data of patients, this may be potentially dangeous as AI cannot be trusted with it. There is also the concern of hacking into the system
Distrupting existing workflows- introduction of AI into a surgical setting will result in changes in the surgical team dynamics and result in surgeons needing additional training.
AI in robotic surgery has shown success in a variety of different areas including supporting cornary artery bypass grafting to brain tumour removal and it has a great potential in healthcare for the future.