HANN: A Hybrid Model for Liver Syndrome Classification by Feature Assortment Optimization

Area of Science:

• Medical Informatics
• Artificial Intelligence in Medicine
• Computational Biology

Background:

• Early disease detection is crucial for effective medical treatment and patient prognosis.
• Liver diseases, including Hepatitis B and C, are increasingly prevalent, necessitating accurate diagnostic tools.
• Current automated liver disease classification systems often lack the accuracy of traditional methods like surgical biopsy.

Purpose of the Study:

• To propose a novel hybrid model for accurate liver syndrome classification and early diagnosis.
• To enhance the accuracy of automated medical diagnosis systems for liver disorders.
• To develop an intelligent system for improved prognosis of liver diseases.

Main Methods:

• A hybrid model combining M-PSO (Multi-Particle Swarm Optimization) for feature selection and M-ANN (Multi-Layer Artificial Neural Network) for classification was developed.
• Patient medical data was analyzed to classify the possibility of disease existence.
• The Spark tool was utilized for examining and evaluating the algorithm's performance.

Main Results:

• The proposed hybrid model demonstrated significantly improved accuracy in liver disease classification compared to existing algorithms.
• The M-PSO feature selection effectively identified relevant input variables for classification.
• The M-ANN algorithm provided reliable classification of liver syndromes based on patient data.

Conclusions:

• The developed hybrid approach offers a more accurate and efficient method for liver disease classification.
• This intelligent system has the potential to aid clinicians in early diagnosis and prognosis of liver disorders.
• The study highlights the effectiveness of integrating advanced data mining techniques for medical data analysis.