We use cookies so we can provide you with the best online experience. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue. Accept and close
Send to Print
Friday. 29 March 2024
Print Page(s)

Models for measuring biological signals to assess risk

This is an excerpt from Low Back Disorders, Third Edition With Web Resource by Stuart McGill.


Biological Signal - Driven Model Approaches


The final approach for risk assessment is to measure biological signals from each subject to capture the unique ways people perform their jobs and then use sophisticated anatomical, biomechanical, and physiological relationships to assign forces to the tissues.


Marras Model and McGill Model

The Marras model (Marras and Sommerich, 1991a, 1991b) measures electromyograms (EMGs) from several muscles and, using known physiological relationships, assigns forces to the muscles during virtually any industrial task. As noted earlier, this approach revealed powerful evidence linking the physical demands of specific occupational tasks with the incidence of LBD.


The McGill model (see chapter 1) uses the same philosophical approach to assign forces to the muscles, but attempts to include the highest level of anatomical accuracy possible. For example, it also measures three-dimensional spine curvature to assess forces to the passive tissues, including the intervertebral discs and the ligaments. By assigning forces to muscles and passive tissues throughout the full range of spine motion, it captures the ways people perform their jobs and even how they change with repetitions of the same jobs. The obvious liability of the approach is its enormous computational requirements, postprocessing of data, and difficulty collecting such comprehensive data from a wide number of workers in the field.


The question of the validity of this type of model must be addressed along with other models based on the biological approach. Some have argued that because these models contain known biomechanical and physiological relationships, they contain a certain amount of content validity. Moreover, both the Marras model and the McGill model have been quite successful in estimating the passive tissue and muscle forces that sum together to produce flexion and extension, lateral bend, and axial twisting moments. (These moments have been well predicted, with the exception of the twisting moment.) In summary, if twisting is not a dominant moment of force in a particular job, these models appear to predict accurate distributions of forces among the support tissues.


EMG-Assisted Optimization Approach

Perhaps the current pinnacle of model evolution is a hybrid modeling approach known as EMG-assisted optimization (developed by Cholewicki and McGill, 1994) with stability analysis and tissue load prediction (Cholewicki and McGill, 1996, was an early example, and Ikeda and McGill, 2012, is a more recent example). This approach exploits the asset of the biological EMG approach to distribute loads among the tissues based on the biological behavior of the subject. It then uses optimization to fine-tune the balancing of joint torques about several low back joints. The optimization takes the biologically predicted forces and adjusts muscle forces the minimal amount possible to satisfy the three-dimensional moment axes at every joint over the length of the lumbar spine. Once the three-dimensional moments have been assigned and tissue forces determined, an analysis of spine stability is performed via comparison of the joules of work imposed on the spine through perturbation with the joules of potential energy existing in the stiffened column.


This most highly evolved of spine models provides the biomechanist or ergonomist with insight into injury mechanisms caused by instability (as witnessed by Cholewicki and McGill, 1992) such as occurs while picking up a very light load from the floor. Even though patients have reported back injury from incidents such as bending over to tie a shoe, previous modeling approaches were sensitive only to tissue damage and injury scenarios produced from large loads and moments. Now a biomechanical explanation is available to explain the subsequent tissue damage, and a method is available to detect the risk of its happening. Although the routine use of this type of sophisticated model by ergonomists is not feasible, it is useful (for trained scientists) for analyzing individual workers and identifying those who are at elevated risk of injury because of faulty personal motor patterns.


Simple or Complex Models?

In summary, the complex models provide a tool to investigate the mechanisms of injury and the effects of technique during material handling on the risk of injury. The most complex and most highly evolved models provide insight into how injury occurs with all types of heavy and light loads. On the other hand, the simpler models, although sacrificing accuracy, can be powerful tools for the routine surveillance of physical demands in the workplace, but they must be wisely interpreted in each case given their limitations and constraints. Biomechanics and ergonomics require the full continuum of models. The choice of which one to use depends on the issue in question.



Website Page URL (Link) Reference:

http://www.humankinetics.com/excerpts/excerpts/models-for-measuring-biological-signals-to-assess-risk?

© 2013 Human Kinetics, Inc. All Rights Reserved.

Return to article