Artificial Intelligence in Healthcare

1. Introduction

The giant progress in artificial intelligence (AI) development has fueled a rapid rise in the applications of various domains over the last decade. Why AI is a big deal? Brynjolfsson & Mcafee (2017) provide two reasons. First, via AI, we can automate many tasks. Second, AI is an excellent learner. Some interesting AI applications have been identified as in Table 1 (Brynjolfsson & Mcafee, 2017; Jordan & Mitchell, 2015). For example, we can take drug chemical properties as the input to the AI applications and the AI applications can produce treatment efficacy as the output. And this type of application can be used in pharmacy R&D.

AI was considered to have the potential to exploit meaningful relationships in a data set and the relationships can be used in the diagnosis, treatment and predicting outcome in many clinical scenarios (Ramesh et al., 2004). AI in healthcare has two main branches: virtual and physical (Hamet & Tremblay, 2017). The virtual branch includes informatics approaches from machine learning to control of health management systems; and the physical branch is represented by robots assist the elderly patient or the attending surgeon.

The use of AI in healthcare has become quite popular, including medical images (Ker et al., 2018), the detection of drug interactions (Han et al., 2022), the identification of high-risk patients (Beaulieu-Jones et al., 2021), etc. Recently, the performance of ChatGPT by Open AI has recently shown amazing performance and attracted many studies.