As the health industry continues to collect more data and the development of Artificial Intelligence continues to reach new heights, the potential collaboration between the two becomes more tempting. Leveraging the power of AI and the sheer amount of data within health could revolutionize the health industry that is known today. However, it is imperative that the ethics of such innovative solutions are considered as while the potential for AI in medicine is astronomical, the potential pitfalls are treacherous. In order to ensure the ethical use of Machine Learning and AI in medicine, a recognized set of ethical guidelines must be put into place for the development of models. AI experts and health professionals alike must work together to consider the ethical quandaries of the usage of AI in medicine and develop methods and guidelines to mitigate them. We have surveyed the current usage and ethical concerns of AI in medicine and the state of ethical machine learning in Computer Science. Through this survey it has become clear that there is a need for a comprehensive framework for ethical machine learning development with applications in medicine. While there is work being done for each stage of the development of machine learning in medicine in an attempt to create ethical models, there are none that cover all of the stages nor are they covering more than a few ethical issues. There is a need for an interdisciplinary, comprehensive framework combining the best of AI Impact Assessments (AIAs), quantitative and qualitative metrics, checklists, and the numerous debates and discussions on the ethics of AI in medicine.
Article ID: 2023GL2
Publisher: Canadian Artificial Intelligence Association