Assessment of heart disease using fuzzy classification techniques

Area of Science:

• Cardiology
• Data Science
• Medical Informatics

Background:

• Accurate classification of cardiac conditions like ischemic cardiopathy, valvular heart disease, and arterial hypertension is crucial.
• Utilizing patient data, including ECHO, stress tests, age, and weight, aids in diagnosis and treatment.
• Fuzzy clustering offers a flexible approach to analyze complex medical datasets.

Purpose of the Study:

• To classify cardiac patients into distinct groups based on multiple health indicators.
• To explore the utility of various fuzzy clustering algorithms for medical data analysis.
• To identify the most influential patient characteristics contributing to observed group similarities and differences.

Main Methods:

• Application of hierarchical fuzzy clustering, hierarchical and horizontal fuzzy characteristics clustering.
• Introduction and use of a novel fuzzy hierarchical cross-classification technique.
• Analysis of 19 patient descriptors, including ECHO data, effort testing results, age, and weight.

Main Results:

• Fuzzy characteristics clustering effectively partitions characteristics, aiding in similarity assessment and selection of essential descriptors.
• Fuzzy hierarchical cross-classification yields partitions of both patients and their characteristics.
• The cross-classification method successfully highlights characteristics responsible for patient group distinctions.

Conclusions:

• Fuzzy clustering algorithms, particularly the novel cross-classification technique, are effective tools for analyzing and classifying cardiac patients.
• This approach facilitates the identification of key patient characteristics driving disease classification.
• The findings support the use of advanced data analysis methods in understanding cardiovascular diseases.