Deep into Laboratory: An Artificial Intelligence Approach to Recommend Laboratory Tests.

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Clinical Decision Support Systems

Background:

• Inappropriate laboratory test ordering leads to suboptimal patient care and increased healthcare expenditures.
• Physician workflow is burdened by manual test selection, limiting time for patient treatment.
• Automated systems can enhance test selection accuracy and efficiency.

Purpose of the Study:

• To develop and evaluate a deep learning-based automated system for recommending appropriate laboratory tests.
• To improve the accuracy and efficiency of laboratory test ordering in clinical practice.

Main Methods:

• Retrospective data collection from the National Health Insurance database (2013).
• Inclusion of 1,463,837 prescriptions from 530,050 unique patients.
• Development and validation of a deep learning model to predict appropriate laboratory tests.

Main Results:

• The deep learning model achieved high performance with AUROC micro = 0.98 and AUROC macro = 0.94.
• The model demonstrated accuracy and efficiency in identifying appropriate laboratory tests.
• The system has the potential to reduce under- and over-utilization of laboratory tests.

Conclusions:

• Deep learning models can accurately and efficiently recommend laboratory tests.
• Integration of this automated system into clinical workflows can optimize laboratory test utilization.
• This approach can improve patient care and reduce financial burdens in healthcare.