Machine-Learning-Powered Information Systems: A Systematic Literature Review for Developing Multi-Objective Healthcare Management

Abstract

The incorporation of machine learning (ML) into healthcare information systems (IS) has transformed multi-objective healthcare management by improving patient monitoring, diagnostic accuracy, and treatment optimization. Notwithstanding its revolutionizing capacity, the area lacks a systematic understanding of how these models are divided and analyzed, leaving gaps in normalization and benchmarking. The present research usually overlooks holistic models for comparing ML-enabled ISs, significantly considering pivotal function criteria like accuracy, precision, sensitivity, and specificity. To address these gaps, we conducted a broad exploration of 306 state-of-the-art papers to present a novel taxonomy of ML-enabled IS for multi-objective healthcare management. We categorized these studies into six key areas, namely diagnostic systems, treatment-planning systems, patient monitoring systems, resource allocation systems, preventive healthcare systems, and hybrid systems. Each category was analyzed depending on significant variables, uncovering that adaptability is the most effective parameter throughout all models. In addition, the majority of papers were published in 2022 and 2023, with MDPI as the leading publisher and Python as the most prevalent programming language. This extensive synthesis not only bridges the present gaps but also proposes actionable insights for improving ML-powered IS in healthcare management.

Ähnliche Arbeiten