Prediction of PCOS and Mental Health Using Fuzzy Inference and SVM

Area of Science:

• Endocrinology and Reproductive Health
• Computational Health Informatics
• Mental Health Research

Background:

• Polycystic Ovarian Syndrome (PCOS) is a prevalent hormonal disorder affecting women of reproductive age.
• Current PCOS detection methods often lack integration with mental health assessments.
• Addressing the comorbidity of PCOS and mental health issues is crucial for comprehensive patient care.

Purpose of the Study:

• To develop and evaluate an automated model for early detection and prediction of PCOS.
• To integrate the assessment of associated mental health issues within the PCOS detection framework.
• To explore the utility of fuzzy logic approaches in modeling the linguistic nature of PCOS symptoms and diagnosis.

Main Methods:

• An automated early detection and prediction model was developed.
• The Fuzzy Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) was employed to handle linguistic symptom data.
• Fuzzy TOPSIS performance was compared against the Support Vector Machines (SVM) algorithm using a local dataset.

Main Results:

• The Fuzzy TOPSIS method achieved a high accuracy of 98.20% in detecting PCOS and associated mental health issues.
• Support Vector Machines (SVM) achieved an accuracy of 94.01% on the same dataset.
• The Fuzzy TOPSIS method demonstrated superior performance compared to SVM.

Conclusions:

• The developed automated model, particularly using Fuzzy TOPSIS, shows significant potential for accurate early detection of PCOS and related mental health conditions.
• Integrating psychological well-being evaluation into PCOS treatment protocols is recommended.
• Early and integrated detection and treatment can facilitate timely preventive measures and improve women's overall health outcomes.