Navigation through a floor maze has been used to assess cognition of individuals. In this paper, we analyze novel performance outcomes in gait and balance during navigation in Virtual Reality Floor Mazes (VRFM). The unique contributions of this paper are: 1) We analyze changes in the gait and balance as subjects navigate through the mazes of different difficulty. These performance metrics are more discriminatory when compared to completion time, which is used frequently in the literature
2) Walking steps are classified into straight steps, turn steps, and spin steps as mazes contain frequent turns
3) Maze difficulty is defined by the number of decision points. We conducted experiments with ten young healthy subjects across three conditions: 1) Control Mazes (CM) wherein the path from the start to the goal was displayed
2) Easy Mazes (EM) which contained a maximum of two decision points
3) Hard Mazes (HM) which contained more than two decision points. The results showed that in hard mazes, subjects took smaller and slower steps with increased gait variability when compared to control mazes or easy mazes. Spin steps showed an increased mediolateral margin of stability in hard mazes compared to easy mazes. The mediolateral center of mass displacement was smaller in straight steps and turn steps in hard mazes when compared to control mazes. These results provide new performance metrics to evaluate navigation in floor mazes. These performance metrics describe how the spatiotemporal parameters of gait change in mazes of different difficulty as opposed to completion time which is a cumulative measure of gait performance.