Formal description of disease courses

Area of Science:

• Medical Informatics
• Clinical Data Analysis
• Computational Linguistics

Background:

• Patient case analysis is a daily, crucial task for clinicians.
• The complexity of patient cases necessitates detailed documentation and discussion, as seen in clinicopathological conferences and the MEDLINE database (over 70,000 cases).
• Current methods for analyzing patient cases against medical literature are limited.

Purpose of the Study:

• To introduce a novel technique, Disease Course Graphs (DCGLs), for modeling disease progression described in medical literature.
• To enable advanced computerised matching between generic disease course descriptions and individual patient case descriptions.
• To enhance the capabilities of computerised patient case analysis.

Main Methods:

• Developing a method to represent disease course descriptions using a graph-based model (DCGL).
• Implementing algorithms for matching DCGLs with patient case data.
• Utilizing medical literature as the source for disease course modeling.

Main Results:

• The proposed DCGL technique effectively models disease course descriptions from medical literature.
• DCGL facilitates advanced computerised matching between literature-based disease courses and individual patient cases.
• This represents a fundamental improvement in computerised patient case analysis.

Conclusions:

• Disease Course Graphs (DCGLs) offer a powerful new approach to analyzing patient cases.
• The ability to match literature-derived disease courses with patient data significantly aids clinical decision-making.
• This technique advances the field of computerised patient case analysis.