Health Informatics via Machine Learning for the Clinical Management of Patients

Area of Science:

• Health Informatics
• Machine Learning
• Clinical Practice

Background:

• Machine learning (ML) methods are increasingly integrated into health informatics systems.
• The impact of these systems on clinical management and practice requires thorough evaluation.

Purpose of the Study:

• To review the impact of health informatics systems utilizing machine learning on patient clinical management.
• To identify advancements in ML methodology within health informatics.

Main Methods:

• Literature review of studies published between 2010-2015.
• Searches conducted in prominent databases like Pubmed, IEEE xplore, and INSPEC.
• Selection of representative examples of health informatics systems advancing ML methodology.

Main Results:

• Significant research activity exists in ML-driven health informatics.
• Few systems demonstrably impact clinical management or practice.
• Methodological advancements are evident, often through collaboration between data scientists and clinicians.

Conclusions:

• Health informatics systems employing machine learning are nascent.
• Large-scale translation of these systems into routine clinical management is yet to be achieved.