Systematic review on machine learning applications in Paralympic sports: current practice and future research

Area of Science:

• Sports Science
• Biomechanical Engineering
• Artificial Intelligence

Background:

• Machine learning (ML) is increasingly applied across various scientific domains.
• Paralympic and adaptive sports present unique challenges and opportunities for technological integration.
• Performance analysis and injury prevention are critical areas for athlete development.

Purpose of the Study:

• To review the applications of ML in Paralympic and adaptive sports.
• To identify challenges hindering ML implementation in these sports.
• To explore ML's role in performance profiling, injury prediction, movement analysis, and assistive technology.

Main Methods:

• A systematic literature search was conducted across six major electronic databases.
• Seventeen relevant studies applying ML in Paralympic and adaptive sports were identified and included.
• Study quality was assessed using the Prediction model Risk of Bias Assessment Tool (PROBAST).

Main Results:

• ML applications were identified in performance profiling, classification, movement analysis, and assistive technology.
• Advancements were noted in para-swimming and wheelchair sports tracking.
• Key challenges include small sample sizes, data heterogeneity, and the need for domain-specific ML adaptations.

Conclusions:

• ML has strong potential to enhance performance analysis and injury prevention in Paralympic sports.
• Current ML models require further adaptation for diverse impairments, equipment, and movement strategies.
• Future research should prioritize robust, interpretable, and adaptable ML models with interdisciplinary collaboration.