Machine learning-based clinical decision support using laboratory data.

Area of Science:

• Laboratory Medicine
• Artificial Intelligence
• Machine Learning

Background:

• Artificial intelligence (AI) and machine learning (ML) are increasingly integral to modern healthcare.
• These technologies are transforming clinical laboratory workflows and patient care.

Purpose of the Study:

• To review the development of ML models in laboratory medicine.
• To examine their impact on clinical laboratory workflow and patient outcomes.
• To discuss the challenges and future directions of ML integration.

Main Methods:

• Summarized the development process of ML models, including data collection, cleansing, feature engineering, and optimization.
• Reviewed the application of ML in Clinical Decision Support Systems (CDSS) for test result interpretation.
• Discussed the integration of ML across pre-analytical, analytical, and post-analytical laboratory phases.

Main Results:

• ML models, including those from automated ML tools, streamline laboratory processes and improve efficiency.
• CDSS utilizing ML aid healthcare professionals in interpreting test results, enhancing decision-making.
• ML integration across all lab phases offers significant potential despite challenges.

Conclusions:

• ML-based CDSS can substantially improve clinical decision-making in healthcare.
• Successful adoption requires addressing model uncertainties, black-box issues, and data acquisition challenges.
• Collaboration, hybrid intelligence, and rigorous validation are crucial for effective implementation.