Systematic Review of Artificial Intelligence Applications in Clinical Trials for Central Nervous System Injuries

Artificial intelligence (AI) has emerged as a promising tool for diagnosing, managing, and treating the injuries of the central nervous system (CNS). The purpose of this study was to evaluate the AI-driven approaches in clinical trials for CNS diseases.

A systematic search of the ClinicalTrials.gov, PubMed, Ovid, and Web of Science databases was conducted to identify interventional trials focusing on CNS injuries. Only interventional studies investigating AI applications in CNS injuries were included, while those targeting neurodegenerative diseases were excluded. Data extraction was performed using a self-designed form.

A total of 51 AI-driven clinical trials for CNS injuries were identified. Most trials focused on screening, diagnosis, monitoring, supportive care, prevention, and treatment, and were primarily conducted in China, the USA, and Europe. The targeted conditions included stroke and its sequelae, traumatic brain injury, and spinal cord injury. All trials employed AI-based tools supported by diverse algorithms, such as convolutional neural networks (CNN), extreme gradient boosting (XGBoost), and K-nearest neighbors (KNN). However, only 21.6% (11/51) of the trials reported outcome data, with 10 demonstrating functional improvements, mainly in motor, swallowing, and neurological performance. Notably, 25.5% of the trials incorporated patient-reported outcome measures (PROMs).

This study demonstrates the growing application of AI in CNS injury management, particularly in diagnosis and treatment. However, the limited reporting of outcomes and underuse of PROMs suggest that most interventions remain in early stages of clinical translation. Standardized trial designs, patient-centered measures, and rigorous performance validation will be essential to ensure the meaningful integration of AI into clinical practice.

AI-based clinical trials for CNS injuries are on the rise, with a focus on diagnosis and treatment. Future trials should prioritize standardized designs, integration of PROMs, and comprehensive performance metrics to ensure clinically meaningful evaluation of AI interventions.