Implementing AI in clinical practice

Question 1

What is the biggest problem that slows down the implementation of AI into clinical practice?

Answer 1

The lack of research about AI in the clinic. Currently, a lot of research is being done on newly developed tools for medical data with the use of machine learning techniques. However, there is a lack of prospective studies and clinical trials. Thus, the effects of AI on a long term are lacking.

Question 2

What is another way to study the effectiveness of AI in the clinic?

Answer 2

Qualitative research, i.e. interviews and focus groups.

Question 3

What are obstacles for the implementation of AI in the clinic?

Answer 3

• Lack of trust
• Privacy
• Liability
• Adversarial attack (machine learning algorithms are trained on specific datasets (i.e. for specific patient populations). When the algorithm is then used for a different dataset, it produces unreliable outcomes)
• Medical staff hesistant to use a new tool (it is a habit, medical staff is used to doing things a certain way)

Question 4