Integrating intramuscular fat radiomics with hamstrings-to-quadriceps structure and function ratios to predict future...

Area of Science:

• Sports Medicine and Imaging Analysis
• Biomechanical Injury Prediction

Background:

• Hamstring strain injuries (HSI) are common in athletes, impacting performance and career longevity.
• Predictive biomarkers for HSI are crucial for developing effective injury prevention strategies.
• Magnetic resonance imaging (MRI)-based radiomics offers a non-invasive method to analyze tissue characteristics.

Purpose of the Study:

• To investigate if radiomic features from thigh intramuscular fat (IMF) can predict future HSI.
• To determine if combining radiomics with muscle imbalance and injury profile data enhances HSI prediction accuracy.

Main Methods:

• Prospective, longitudinal study analyzing IDEAL MRI scans of 93 professional American football players.
• Extraction of radiomic features from intramuscular fat (IMF) in hamstring and quadriceps muscles.
• Utilized K-nearest neighbor classifier to evaluate prediction models including radiomics, muscle imbalance (Mb), and injury profile (Mi) features.

Main Results:

• The integrated model combining hamstring IMF radiomics (Mr) with muscle imbalance (Mb) and injury profile (Mi) data achieved the highest prediction accuracy (AUC=0.79).
• This integrated model significantly outperformed models using only radiomics (Mr or quadriceps IMF radiomics), muscle imbalance, or injury profile features individually.
• Radiomics of hamstring IMF alone showed moderate predictive capability (AUC=0.69).

Conclusions:

• Radiomic features from hamstring IMF, when integrated with muscle imbalance and injury profile data, can effectively predict future HSI.
• This multimodal approach offers a promising strategy for identifying athletes at risk of HSI.
• Further validation in larger, diverse athletic populations is warranted to broaden applicability across sexes and participation levels.