Advancing Laboratory Medicine Practice With Machine Learning: Swift yet Exact.

Area of Science:

• Medical Informatics
• Clinical Laboratory Science
• Artificial Intelligence

Background:

• Machine learning (ML) is increasingly utilized for data analysis and prediction across diverse scientific fields.
• Laboratory medicine presents a growing area for ML application, impacting diagnostics and operational efficiency.
• A comprehensive review is needed to understand the current landscape of ML in laboratory medicine.

Purpose of the Study:

• To systematically review and categorize the applications of machine learning in laboratory medicine.
• To identify trends in ML model usage, data types, and research objectives within the field.
• To evaluate the impact of ML on laboratory efficiency and the expansion of clinical laboratory roles.

Main Methods:

• A systematic literature search was conducted on PubMed for articles published between February 2014 and March 2024.
• 144 relevant articles were selected from an initial pool of 779.
• Data extraction and categorization included ML models, specimen types, data types, research objectives, evaluation metrics, and sample sizes, visualized using Sankey diagrams and pie charts.

Main Results:

• Most ML applications in laboratory medicine aim to enhance efficiency via automation and broaden the scope of clinical laboratory services.
• Convolutional neural networks, multilayer perceptrons, and tree-based models are the most frequently employed ML algorithms.
• Model selection is predominantly driven by the nature of the input data.

Conclusions:

• Machine learning is poised to become a more significant tool in laboratory medicine, facilitating expanded research capabilities.
• There is a need to enhance expertise in ML applications to fully leverage its potential within laboratory medicine.
• Continued evolution of ML technology will likely drive further integration and innovation in clinical laboratories.