A Rule-Based System for Personalized Yoga Therapy Management
A Rule-Based System for Personalized Yoga Therapy Management
Yoga, a timeless holistic discipline rooted in ancient Indian philosophy, has emerged as a global phenomenon for enhancing human well-being. Originally designed as a path toward self-realization, yoga today is widely embraced for its therapeutic potential in improving both physical and mental health. Despite the growing evidence supporting its benefits, especially in clinical settings, its adoption remains hindered by challenges, including the lack of a structured yet personalized practice framework. Personalized yoga practices often lead to subjective practice selection and variability among experts.
This project presents a rule-based decision support system designed to generate personalized and evidence-informed yoga therapy modules for individuals with one or multiple health conditions. The system integrates disease-specific yoga modules curated through empirical evidence, including randomized controlled trials and expert-validated content validity ratio scores, along with documented contraindications.
The system operates through a dual-panel architecture: an admin panel for structured data management and a user panel for therapeutic personalization. The admin panel stores disease-wise yoga modules, wherein each practice is mapped to predefined therapeutic categories (including preparatory practices, yogāsana, prāņāyāma, meditation, lifestyle modifications, and yogic diet) as well as categories designated under the Panchakosha framework (Annamaya, Pranamaya, Manomaya, Vijnanamaya, and Anandamaya). Each practice is embedded with evidence parameters such as the number of supporting randomized controlled trials and content validity ratio scores, along with disease-specific contraindication data.
The user panel, intended for yoga consultants, enables users to select a primary disease as well as multiple comorbid conditions, arranged according to clinical severity. Users allocate weighted priority scores (total equals one hundred) reflecting disease severity. Based on these inputs, the system systematically computes category-wise practice distribution. Practices are ranked using a hierarchical priority algorithm based on number of positive randomized controlled trials, repetition across selected disease modules, and content validity ratio scores.
The system also performs automated contraindication filtering, removing any practice contraindicated for even a single selected condition. Fractional allocations arising from weighted computations are resolved using a severity-priority rule, ensuring clinically dominant conditions receive precedence. The final output is presented in an Integrated Yoga Therapy format, structured kosha-wise, thereby translating computational results into a therapeutically coherent clinical module.
This system introduces a novel technical method for evidence-based yoga therapy personalization while simultaneously standardizing the clinical decision-making process, enabling safe, transparent, and reproducible yoga therapy prescriptions for multimorbidity management.