Automated event detection algorithm for two squatting protocols

Area of Science:

• Biomechanics
• Motion Analysis
• Human Movement Science

Background:

• Accurate identification of squatting events is crucial for motion analysis in biomechanics.
• Automated event detection enhances consistency and efficiency in processing squatting trials.
• Previous methods lacked standardized criteria for event identification.

Purpose of the Study:

• To develop and validate criteria for automatically identifying key squatting events.
• To apply these criteria to two distinct squatting protocols.
• To assess the effectiveness in both pathological and typically developing populations.

Main Methods:

• Developed event identification criteria using sagittal plane knee and vertical center of mass velocities.
• Compared absolute versus relative thresholds for peak knee velocity.
• Implemented criteria into an automatic event detection algorithm.
• Tested on two squatting protocols (hold squat, traditional squat) in hip dysplasia patients and typically developing individuals.

Main Results:

• The automated event detection algorithm successfully identified events in 94% of all trials.
• A relative threshold algorithm proved effective for event identification.
• Manual event placement, when necessary, demonstrated perfect inter-rater reliability.

Conclusions:

• The developed criteria provide a highly successful and reliable method for automatic squatting event detection.
• This approach is effective across different squatting protocols and participant groups.
• The automated system improves consistency and reduces processing time in biomechanical analysis.