Systematic Review of Data Mining Applications in Patient-Centered Mobile-Based Information Systems

Area of Science:

• Health Informatics
• Mobile Health (mHealth)
• Data Mining Applications

Background:

• Smartphones are increasingly vital in patient-centered healthcare.
• Data mining techniques are leveraged to personalize mobile health applications.
• This review examines data mining in patient-centered mobile information systems.

Purpose of the Study:

• To review current literature on data mining applications in patient-centered mobile health.
• To identify how data mining improves mobile apps for patient needs.
• To assess the role of data mining in mobile-based healthcare systems.

Main Methods:

• Systematic literature search of PubMed, Scopus, and Web of Science (2014-2016).
• Screening of 226 records at title/abstract level.
• Full-text review of 92 papers, with 30 included in the final analysis.

Main Results:

• Data mining in mobile health apps primarily supports data analysis for monitoring (n=27), early diagnosis, classification/prediction, and risk calculation.
• Fewer applications focus on data collection (n=3) or providing recommendations (n=2).
• Support vector machine was frequently used, while decision trees showed superior performance for patient self-management apps.

Conclusions:

• Data mining features in mobile apps can improve patient care through case detection, prediction, risk estimation, and data collection, especially during self-management.
• Advanced methods like artificial neural networks and hybrid approaches can enhance personalized recommendations, education, and alerts.
• Integrating data mining into mobile health fosters more patient-centered care and improved health outcomes.