Predicting Temperament and Cardiac Strength from Persian Medicine Pulsology Using Artificial Neural Networks: An Algorithmic and Sensitivity Analysis
محل انتشار: نشریه علم داده و مدل سازی، دوره: 3، شماره: 2
سال انتشار: 1404
نوع سند: مقاله ژورنالی
زبان: انگلیسی
مشاهده: 30
فایل این مقاله در 22 صفحه با فرمت PDF قابل دریافت می باشد
- صدور گواهی نمایه سازی
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
JR_JCSM-3-2_006
تاریخ نمایه سازی: 22 آبان 1405
چکیده مقاله:
Persian Medicine (PM) pulsology is a central diagnostic modality for assessingan individual’s temperamental state and cardiac strength; however, its practicalapplication is inherently subjective and dependent on practitioner expertise. Toenhance objectivity and reproducibility in PM diagnostics, this study proposesan artificial neural network (ANN) framework for the quantitative estimation offour core temperamental qualities—warmness, coldness, wetness, and dryness—together with cardiac strength, based solely on measurable pulse characteristicsClinical data were collected from ۶۹ individuals, comprising ۱۱ pulse-derived featuresand gender as inputs, with six corresponding diagnostic outputs. A multilayerperceptron (MLP) architecture was trained and optimized using two learningalgorithms: Levenberg–Marquardt (LM) and Scaled Conjugate Gradient (SCG).Comparative evaluation demonstrated the superiority of the SCG-trained network,with an optimal configuration of ۱۲ hidden neurons. This model achieved a testaccuracy of ۹۶.۴۳% and a low mean squared error (MSE) of ۰.۰۱۳۹. Notably, cardiacstrength was predicted with ۱۰۰% accuracy, while temperamental qualitieswere classified with accuracies ranging from ۷۱.۴۳% to ۹۲.۸۶%. To enhance interpretability,a sensitivity analysis was conducted, revealing pulse strength and pulsefrequency as the most influential predictors across multiple diagnostic outputs.The proposed ANN-based system provides a stable and objective computationalsurrogate for traditional PM pulsology. It offers practical utility for practitionertraining, supports diagnostic standardization, and establishes a methodologicalfoundation for future integrative and intelligent diagnostic platforms in PersianMedicine.
کلیدواژه ها:
Persian Medicine (PM) ، Artificial Neural Network (ANN) ، Pulse Diagnosis ، Temperamental ، Scaled Conjugate Gradient (SCG)