Increased Pre-Activation and Co-Contraction in ACL-Reconstructed Athletes: Insights from AI-based EMG Analysis.

Area of Science:

• Sports Medicine
• Biomechanics
• Rehabilitation Science

Background:

• Anterior Cruciate Ligament (ACL) tears disrupt knee neuromuscular control, impacting muscle activity during movement.
• Evaluating muscle activation patterns is crucial for return-to-sport assessments after ACL Reconstruction (ACL-R).
• Artificial intelligence (AI) offers potential for precise analysis of muscle activity data.

Purpose of the Study:

• To quantify knee muscle pre-activation and co-contraction indexes in athletes with and without ACL-R during landing tasks.
• To utilize an AI approach for reliable assessment of neuromuscular control.
• To compare AI-derived parameters with traditional methods.

Main Methods:

• Electromyography (EMG) signals were recorded from four knee muscles in 11 ACL-R athletes and 18 control athletes during single-leg hop and cross drop landings.
• A deep learning-based muscle activity detector (LSTM-MAD) was employed to compute muscle pre-activation onsets and timing-based co-contraction indexes.
• Amplitude-based co-contraction indexes were also calculated for comparison.

Main Results:

• The AI method accurately estimated muscle pre-activation onset with an error under 23ms.
• ACL-R athletes exhibited significantly greater Biceps Femoris (BF) and SemiTendinosus (ST) pre-activations and longer co-contraction durations during single-leg hops compared to controls.
• ACL-R athletes also showed greater BF pre-activation and longer BF/ST co-contractions during cross drop landings.

Conclusions:

• ACL-R athletes demonstrate persistent alterations in neuromuscular control, specifically increased pre-activation and co-contraction of BF and ST muscles during landing.
• AI-based assessment of neuromuscular control can enhance the effectiveness of rehabilitation.
• This approach can support safer and more informed return-to-sport decisions for athletes post-ACL-R.