Automated Orientation & Mobility Training in Virtual Reality for Low Vision Rehabilitation

The long-term objective of the research is to use Virtual Reality-based Intelligent Orientation & Mobility Specialists (VR-IOMSs) to teach a large portion of the orientation & mobility (O&M) curriculum to low vision (LV) patients. O&M, the ability to orient oneself and to travel independently, safely and efficiently in any environment, is the foundation of independent living and high quality of life. O&M rehabilitation is the only proven intervention that restores mobility lost to LV. Current O&M rehabilitation is conducted by certified O&M specialists (COMS) in real streets. The shortage and uneven distribution of COMS, the mobility difficulty of LV patients, the long training needed to regain independent travel, the non-reimbursable cost of COMS time and the low income/unemployment status of the LV population make this valuable service inaccessible and unaffordable to many who need it the most. A VR-IOMS, a computer program, combines virtual reality and intelligent tutoring technologies. When VR-IOMS courses are delivered through the internet, LV patients can receive quality O&M training in safe virtual environments at their convenient location and time with little cost. This new intervention can thus reduce the accessibility and affordability barriers to O&M rehabilitation. The proposed research is built on our previous study of teaching O&M skills in virtual streets. It is designed to test the hypothesis that there is no difference in real street O&M performance between LV patients who receive O&M training from VR-IOMSs and those who receive training from human COMSs in real streets. Three specific aims will be achieved. 1) Design VR-IOMSs. Three VR-IOMSs for three important O&M tasks, timing to cross a signalized street, timing to cross an uncontrolled street and learning the outdoor numbering system, will be designed. This involves collecting expert knowledge about teaching these tasks through focus group research, defining intelligent course flow and designing course contents in virtual reality. 2) Implement the VR- IOMSs. This involves building virtual reality simulators, developing virtual street scenarios, authoring the intelligent course, programming interfaces between the intelligent O&M specialists and virtual street scenarios. 3) Validate the VR-IOMSs. A randomized training trial will be conducted to compare VR-IOMS and human COMS O&M training. LV subjects will be randomized to learning the three O&M tasks from the VR-IOMSs, from human COMSs and no training. All subjects will be taken to real streets by COMSs before and after training to evaluate their O&M task performance. Pre- and post-training performance changes of the three groups will be compared to quantify relative training effectiveness. If these VR-IOMSs are proven effective, more will be developed and validated in future research. Clinical deployment of these training agents will restore independent travel to more LV patients. The VR-IOMSs, with their explicitly defined training strategies, objective assessments of learner behavior and quantitative outcome measures, can serve as platforms to advance O&M research. Experience gained here can be used to improve other fields of rehabilitation.