Classification of otoneurological cases according to Bayesian probabilistic models

Area of Science:

• Medical Informatics
• Computational Biology
• Machine Learning in Healthcare

Background:

• Otoneurological diseases present diagnostic challenges, even for specialists.
• Accurate classification is crucial for effective patient management and treatment.

Purpose of the Study:

• To evaluate the efficacy of Bayesian methods for classifying six otoneurological diseases.
• To assess the ability of Bayesian networks to determine attribute dependencies in otoneurological data.

Main Methods:

• Utilized a naive Bayesian probabilistic model and Bayesian networks with various scoring functions.
• Employed a dataset comprising 38 otoneurological attributes.
• Conducted classification using tenfold cross-validation.

Main Results:

• Achieved average sensitivities of 90% and positive predictive values of 92%.
• Reached classification accuracies as high as 97%.
• Demonstrated superior performance compared to previous neural network models.

Conclusions:

• Bayesian methods offer a powerful and efficient tool for classifying otoneurological diseases.
• These methods accurately interpret complex medical data and knowledge.
• Bayesian approaches show significant potential for improving diagnostic accuracy in otoneurology.