Fuzzy AI Using VR for Collision Prevention

Not Applicable

Completed

• Conditions: Injuries
• Interventions: Other: GFT AI training

• Registration Number: NCT03915119
• Lead Sponsor: Children's Hospital Medical Center, Cincinnati

Brief Summary

The purpose of this study is to develop and test a VR training system that integrates GFT AI with virtual obstacle scenarios that, when compared to a sham-VR training system, is hypothesized to increase neuromechanical and perceptual-motor fitness, decrease collision frequency and impact forces for soccer athletes, during a single training session and also when assessed at approximately 1 week and 1 month following training.

Detailed Description

Player collisions cause over 70% of concussion injuries in contact sports, in addition to 50% of lower extremity injuries and 40% of catastrophic knee ligament injuries. The majority of these collisions are unanticipated, and associated with reduced neuromechanical and perceptual-motor fitness underlying an athlete's adaptability to on-field conditions. Thus, training collision anticipation necessitates a method that taps into neuromechanical and perceptual-motor fitness. Virtual reality (VR) is a tool that can target these mechanisms, while providing a safe, well-controlled environment for assessment and training. The current proposal innovates on VR with the integration of genetic fuzzy tree (GFT) artificial intelligence (AI) to drive scenario configuration designed to target modifiable mechanisms and tailored to the individual athlete's performance capabilities, for the optimization of behavior modification and skill transfer. The current study will examine test a GFT AI-driven VR collision anticipation training compared to a sham-VR training system in healthy soccer athletes.

Study Timeline

• Study Start: 2018-01-01
• Study First Submitted: 2018-10-30
• Primary Completion: 2019-02-01
• Study First Submitted QC: 2019-04-11
• Study First Posted: 2019-04-16
• Study Completion: 2019-09-30
• Status Verified: 2020-02
• Last Update Submitted: 2020-02-06
• Last Update Posted: 2020-02-07

Recruitment & Eligibility

• Status: COMPLETED
• Sex: All
• Target Recruitment: 45

Inclusion Criteria

• healthy
• varsity high school or collegiate soccer athlete

Exclusion Criteria

• unable to participate in soccer
• history of congenital or acquired cognitive, ophthalmologic, or neurological disorders including developmental delay, brain tumor, stroke, or known peripheral or central vestibular disorders
• patients who have begun anti-depressant, stimulant or anti-seizure medications for treatment of their symptoms or for other, unrelated reasons within two months of testing will be excluded from testing

Study & Design

• Study Type: INTERVENTIONAL
• Study Design: PARALLEL

Arm && Interventions

Group	Intervention	Description
VR Obstacle group	GFT AI training	Initial Visit: navigates a cluttered array of fixed and moving virtual obstacles to reach a way-point (goal) as fast and efficiently (avoiding obstacles) as possible. In each block, task difficulty (i.e., complexity) will increase linearly, regardless of success or failure (≥ 1 collision before reaching the goal). Second (Training) Visit (randomized into two groups):
Agility Group	GFT AI training	Initial Visit: Completes a soccer ball dribbling agility task in which they must dribble a soccer ball toward an artificial way-point, while avoiding artificial obstacles overlaid onto the real world via a Microsoft Hololens augmented reality display. Second (Training) Visit

Primary Outcome Measures

Name	Time	Method
Pro-saccades test	10 minutes	Pro-saccades will be tested as participants track discrete target motion that will jump randomly by 14, 16, 18, 20, 22, or 24˚ on the screen in a horizontal and vertical direction, at intervals varying pseudo randomly between 1.0 and 2.0 s. The current fixation target will be extinguished at the same time as the next peripheral target appears. The test sequence will take 30 s/trial (2 trials) and all participants will be instructed to follow the targets as quickly and accurately as possible

Secondary Outcome Measures

Name	Time	Method
Self paced saccades	10 minutes	Self-paced saccades will be assessed as the participant glances back and forth as quickly and accurately as possible between two constantly visual targets at ± 15˚ horizontally from one another. This test will take 30 s per trial and the participant will perform 2 trials.
Anti-saccades test	10 minutes	Anti-saccades will be tested as participants are presented discrete targets that jump randomly by 14, 16, 18, 20, 22, or 24˚ on the screen in a horizontal and vertical direction, at intervals varying pseudo randomly between 1.0 and 2.0 s while a cross-hairs is present in the center of the screen. The athletes will be instructed to fixate on the cross-hairs and at the moment each object appears, to saccade in the opposite direction and then return their gaze back to the cross-hairs for the start of the next trial.

Trial Locations

Locations (1)

Cincinanti Childrens Hospital Medical Center; 🇺🇸 Cincinnati, Ohio, United States