Video based lifting technique coding system.

Abstract

Despite automation and improved working conditions, many materials in industry are still handled manually. Among the basic activities involved in manual materials handling, lifting is the one most frequently associated with low-back pain (LBP). Biomechanical analysis techniques have been used to better understand the risk factors associated with manual handling, but because these techniques require specialized equipment, highly trained personnel, and interfere with normal business operations, they are limited in their usefulness. A video based lifting technique analysis system (the VidLiTeCTM System) is presented that provides for quantifiable non-invasive biomechanical analysis of the dynamic features of lifting with high inter-coder reliability and low sensitivity to absolute errors. Analysis of results from a laboratory experiment and from field-collected videotape are described that support the reliability, sensitivity, and accuracy claims of the VidLiTeCTM System. The VidLiTeCTM System allows technicians with minimal training and low-tech equipment (a camcorder) to collect large sets of lifting data without interfering with normal business operations. A reasonably accurate estimate of the peak compressive force on the L5/S1 joint can be made from the data collected. Such a system can be used to collect quantified data on lifting techniques that can be related to LBP reporting.