The influence of training on decision times and errors associated with classifying trunk postures using video-based...

Area of Science:

• Ergonomics and Biomechanics
• Human Factors Engineering
• Occupational Health

Background:

• Video-based posture assessment is crucial for ergonomic evaluations and injury risk prediction.
• Accurate and efficient trunk posture classification is essential for reliable ergonomic assessments.
• The impact of targeted training on classification performance in video-based methods requires further investigation.

Purpose of the Study:

• To investigate how training influences decision times and error rates in video-based trunk posture classification.
• To compare the effectiveness of training on classification performance between amateur and knowledge-based participants.
• To determine the feasibility of training for improving the efficiency of video-based posture analysis in large-scale studies.

Main Methods:

• A three-phase study (pre-training, training, post-training) involving 60 participants (30 amateur, 30 knowledge-based).
• Participants classified static trunk postures from images presented on a computer screen.
• Postures were shown in flexion/extension and lateral bend views at varying distances from category boundaries.

Main Results:

• Both decision time and classification errors decreased as the distance from posture category boundaries increased.
• Amateur analysts showed a greater reduction in decision time per classification (0.79s) compared to knowledge-based analysts (0.60s).
• Training led to significant improvements in classification performance for both groups.

Conclusions:

• Training can effectively reduce decision time and errors in video-based trunk posture classification.
• These findings suggest that training enhances the efficiency and accuracy of ergonomic assessments.
• Optimized training protocols can make video-based posture assessment methods more feasible for large-scale field studies and injury risk assessment.