Beyond the Ball: An AI-Driven Ecosystem for Real-Time Diaphragmatic Coaching and Dual-Mode Lung Function Assessment in Asthma Management

Abstract

Background: Respiratory health management requires accessible tools for both rehabilitation and diagnostic monitoring. Traditional devices like incentive spirometers and peak flow meters are limited by user error and provide incomplete data that patients find difficult to interpret. Objective: This paper presents a conceptual framework for the AI-Driven Pranayama Spiroball, a proposed dual-action biofeedback system designed to integrate rehabilitative inhalation guidance using Pranayama principles with expiratory function measurement through a dedicated spirometric sensor. Material and Methods: The proposed system architecture comprises two functional modes: (1) rehabilitative inhalation using a standard spiroball with AI-guided diaphragmatic breathing coaching via smartphone camera-based abdominal tracking, and (2) diagnostic exhalation using a calibrated flow sensor to measure PEF, FEV1, and FVC. We propose a composite metric, the Lung Efficiency Index (LEI), combining Inspiratory Capacity Score, Expiratory Function Score, and Technique Adherence Score. This paper details the system design, technical specifications, and outlines a validation roadmap. Results: The conceptual framework aims to reduce user error through real-time AI feedback and simplify complex respiratory data into a single interpretable metric (LEI, 0-100 scale). Clinical validation is required to establish efficacy. Conclusions: This design framework addresses identified gaps in respiratory care by proposing unified rehabilitation and monitoring functions. Pilot studies with healthy volunteers and subsequent clinical trials are necessary to validate the system's effectiveness and the LEI's predictive value. A functional prototype is currently under development.