Uniplanar aquatic exercise quantified with inertial sensors and pose estimation

Area of Science:

• Biomechanics
• Sports Technology
• Rehabilitation Engineering

Background:

• Traditional marker-based motion capture for aquatic exercise is expensive and labor-intensive.
• There is a need for accessible, objective methods to quantify human movement in aquatic environments.
• Current technology limits detailed analysis of aquatic exercise performance.

Purpose of the Study:

• To concurrently validate inertial measurement unit (IMU) sensors and markerless computer vision (CV) pose estimation for quantifying uniplanar human movement during aquatic exercise.
• To assess the applicability of IMU and CV methods as alternatives to traditional motion capture.
• To establish the validity of these novel methods for deriving key performance metrics in aquatic settings.

Main Methods:

• Concurrent validity analysis comparing IMU sensors and CV pose estimation against established motion capture systems.
• Quantification of uniplanar aquatic exercise (knee and hip flexion-extension) using derived metrics: range of motion, duration, and angular velocity.
• Intraclass correlation coefficients (ICC) were used to assess agreement between measurement techniques.

Main Results:

• Both IMU sensors and CV pose estimation demonstrated excellent agreement (ICC ≥ 0.94) with traditional methods for all performance metrics.
• High reliability was observed across all derived metrics, including range of motion, duration, and angular velocity.
• The proposed methods proved effective in quantifying uniplanar movements during aquatic exercise.

Conclusions:

• IMU sensors and CV pose estimation offer robust and accessible alternatives for objective exercise quantification in aquatic environments.
• These technologies represent a significant step towards overcoming the limitations of traditional motion capture in specialized settings.
• The findings support the integration of IMU and CV systems for enhanced aquatic exercise analysis and monitoring.