ARGO 2.0: a Hybrid NLP/ML Framework for Diagnosis Standardization

Area of Science:

• Medical Informatics
• Artificial Intelligence in Medicine

Background:

• Diagnosis formulation is subjective and uncertain, necessitating standardization.
• Digital solutions can automate routines to improve diagnostic consistency.

Purpose of the Study:

• To introduce ARGO 2.0, a framework for developing AI-driven decision support systems for diagnosis formulation.
• To demonstrate ARGO 2.0's template-independent and adaptable nature for various medical fields.

Main Methods:

• ARGO 2.0 processes free-text medical reports to create electronic Case Report Forms.
• A hybrid approach combining Natural Language Processing (NLP) and Machine Learning (ML) automates standardized diagnosis suggestions.

Main Results:

• ARGO 2.0 achieved high performance in hemo lympho-pathology with 95.07% Accuracy, 94.85% Precision, 96.31% Recall, and 95.32% F-Score.
• The integrated hybrid strategy outperformed individual NLP and ML components.

Conclusions:

• ARGO 2.0 offers a feasible and effective framework for standardizing medical diagnosis formulation.
• The system's adaptability and high accuracy show promise for improving clinical decision support.