Hey Dr, do you need AI?
Despite the rapid global progress in research and development in innovative medical AI technologies, there are several barriers that hinder its clinical use at a wider scale.
As a healthcare cum digital health professional , it is always exciting and concerning on about AI influx in healthcare systems.
I wanted to dig more and understand the ethical challenges and the legal requirements behind the use of AI in healthcare. This led me to take up this online course from openWHO. The course titled “Ethics and Governance of Artificial Intelligence for Health” provided me with lots of insights on the different aspects of ethics surrounding the use of AI.
View the course contents here: https://openwho.org/courses/ethics-ai/overview
Some of the points below are my learnings from this course.
As a doctor what should you consider before embracing AI in your clinical activity!
Is it necessary to use AI in my scope of services?
Whether the context — scope of use and population — in which the AI technology will be used is appropriate?
Do I have sufficient training to use these tools?
Are there any existing conflict of interests that could cause potential harm?
Whether the use of AI should be justifiable in the context of health care requirements and social needs?
Whether patient autonomy is preserved while using AI technology?
Can I ensure data privacy protection?
Will the tool undermine or bias my decision making skills?
What is the expected error level of the tool?
Have you answered all the above questions? Then answer these extended questions below…
Is the tool safe, accurate and efficient?
Can the decisions of the tool be overlooked?
What is the risk-to-benefit ratio?
Does the tool inculcate the concept of “Humans in the Loop”?
Is the tool transparent?
Is the data used to train the tool and its algorithm known?
Is the data “inclusive”?
Does the tool address data selection bias, stigmatization or discrimination due geographical location, gender, ethnicity, race, health status etc.?
Can the tool safeguard privacy?
Do you have the skills and the resources to regularly challenge and review the tool?
Does the tool consider changes in context over time?
The above list of questions were the ones that were interesting from my view point. From any healthcare provider’s perspective, even though the current landscape glorifies AI, it would be wise for healthcare providers to approach the use of these tools with a critical and thoughtful perspective.
Be wise, play hard but play safe!
Cheers to AI :) The future is bright!
