Machine Learning in Evidence Synthesis Research

In this chapter, we will explore how Systemic Reviews (SR) are traditionally conducted and how the process of arriving at a valuable SR can be made more efficient and less error prone using Machine Learning (ML) techniques. As the integration of ML at the screening stage of SRs has reached the highest level of maturity, we will explain the techniques utilized. We will further describe the extraction process from primary studies supported by ML techniques. The discussion of pitfalls when conducting SRs concludes the chapter, specifically how ML can address bias. Lastly, we address the inherent limitations of artificial intelligence in healthcare with a special emphasis on ML for the use in SRs.