Classification of Post-Anterior Cruciate Ligament Reconstruction Running Dynamics using Non-Traditional Features

A new metric for characterizing limb loading dynamics in post anterior cruciate ligament reconstruction individuals

BACKGROUND

Unresolved neuromuscular deficits often persist in post-anterior cruciate ligament reconstruction (ACLR) individuals manifesting as altered impact and active peak force production during running that can contribute to detrimental limb loading. Elevated impact and active peaks are common in pathological populations indicating a stiffer limb loading strategy. Although impact and active peaks are sensitive to changes in limb loading, to our knowledge, there are no established, standardized measures or cutoff criteria to differentiate between healthy and pathological limb loading. However, prior studies have demonstrated that the ratio between traditional biomechanical measures can be used to successfully establish quantifiable and graphical ranges to delineate between healthy and pathological movement.

RESEARCH QUESTION

Therefore, this study sought to exploit the impact-to-active peak ratio to generate a new, standardized metric to quantify and characterize limb loading dynamics in healthy controls and post-ACLR individuals during running.

METHODS

Twenty-eight post-ACLR individuals and 18 healthy controls performed a running protocol. Impact peak and active peak data were extracted from their strides as they ran at a self-selected speed. A linear regression model was fit to the healthy control data and the models 95 % prediction intervals were used to define a boundary region of healthy limb loading dynamics.

RESULTS

The post-ACLR individuals produced a higher impact-to-active peak ratio than the healthy controls indicating that they adopted a stiffer limb loading strategy. The boundary regions derived from the impact and active peak model successfully classified the healthy controls and post-ACLR individual's limb loading dynamics with an accuracy, sensitivity, and specificity of 89 %, 100 %, and 75 %, respectively.

SIGNIFICANCE

The ability to effectively evaluate limb loading dynamics using impact and active peaks can provide clinicians with a new, non-invasive metric to quantify and characterize healthy and pathological movement in a clinical setting.

A Metric for Identifying Stress Fractures in Runners

PURPOSE

Stress fractures are common overuse running injuries. Individuals with stress fractures exhibit running biomechanics characterized by elevated impact peak and loading rate. While elevated impact peak and loading rate are associated with stress fractures, there are few established metrics used to identify the presence of stress fractures in individuals. Here this study aims to exploit the linear relationship between the impact peak and loading rate to establish a metric to help identify individuals with stress fractures. We hypothesize that the ratio between the impact peak and loading rate will serve as a metric to delineate between healthy controls and those with stress fractures.

METHODS

Fifteen healthy controls and 11 individuals with stress fractures performed a running protocol. A linear regression model fit to the stress fracture impact peak and loading rate data produced a lower 95% confidence limit boundary that served as the demarcation line between the two groups.

RESULTS

Individuals with stress fractures tended to reside above the line with the line accurately classifying 82% of the individuals with stress fractures.

CONCLUSION

The analysis supported the hypothesis and demonstrated how the relationship between impact peak and loading rate can help identify the presence of stress fractures in individuals.Clinical Relevance- The relationship between impact peak and loading rate has the potential to serve as clinically useful metric to identify stress fractures during running.