SmartHMD for Improved Mobility

Not Applicable

Completed

Conditions: Low Vision
Mobility Limitation
Visual Impairment
Navigation, Spatial
Orientation

Interventions: Device: No HMD used
Device: HMD worn but not active
Device: HMD worn and active

Registration Number: NCT03781583

Lead Sponsor: James Weiland

Brief Summary: The National Eye Institute estimated about 3 million people over age 40 in the US had low vision in 2010 and projects an increase to nearly 5 million in 2030 and 9 million in 2050. Current assistive technologies are a patchwork of mostly low-technology aids with limited capabilities that are often difficult to use, and are not widely adopted. This shortfall in capabilities of assistive technology often stems from lack of a user-centered design approach and is a critical barrier to improve the everyday activities of life (EDAL) and the quality of life (QOL) for individuals with low vision.

An intuitive head mounted display (HMD) system on enhancing orientation and mobility (O\&M) and crosswalk navigation, could improve independence, potentially decrease falls, and improve EDAL and QOL. The central hypothesis is that an electronic navigation system incorporating computer vision will enhance O\&M for individuals with low vision. The goal is to develop and validate a smartHMD by incorporating advanced computer vision algorithms and flexible user interfaces that can be precisely tailored to an individual's O\&M need. This project will address the specific question of mobility while the subject crosses a street at a signaled crosswalk. This is a dangerous and difficult task for visually impaired patients and a significant barrier to independent mobility.

Detailed Description: This study was originally intended as a single-arm study. Early results and unanticipated events related to component availability made it clear that multiple configurations of the device would be necessary to produce a functional device. Therefore, the study was split into multiple cohorts representing each iteration of the device and testing process, as follows:

As the trial progressed, the participants recruited were consistently unable to perform tasks without smartHMD. This is due to the recruitment of participants with severe vision loss. Due to this finding, data was only acquired when participants used the smartHMD when active instead of collecting data with and without the system being active.

The participants performed a subset of experiments due to time constraints. Some performed only outdoor trials; some performed only indoor trials; others performed both. The outcome measures depend on which experiments were performed.

After the first test involving the initial three participants, the ODG system became unsupported by the company that produced the device. In response, the study team developed a prototype system (RGBD), which was used in the ODG system's place for the remaining tests. The RGBD was initially used indoors with laptop and subsequently outdoors with mobile hardware and new software.

Recruitment & Eligibility

Status: COMPLETED

Sex: All

Target Recruitment: 21

Inclusion Criteria

Diagnosed with low vision
Self reported difficulty with mobility and finding doors (either indoors or outdoors) and using signalized crosswalks.
Ability to cooperate for tests
Able to participate in all visits

Exclusion Criteria

Unable to use head mounted display technology
Unstable age-related macular degeneration within the past 3 months
Unstable diabetic retinopathy within the past 3 months
Unstable diabetes within the past 3 months
Ocular infection or ocular inflammation in the past 3 months
Ocular trauma within the past 6 months
Intraocular surgery within 6 months
Optical coherence tomography retinal findings of concern to investigator for unstable vision during the study
Women who are pregnant (due to risk of falls and change in gait).
Uncontrolled seizure disorder in the past 6 months
Cerebrovascular accident occurring in the past 6 months
Parkinson disease or neurological condition that limits mobility
Alzheimer disease or other forms of dementia
Conditions of concern to investigator that would confound orientation and mobility, such as severe arthritis, pain that limits ambulatory activities, or orthopedic surgery (e.g., hand, arm, shoulder, knee, or hip surgery within 12 months)

Study & Design

Study Type: INTERVENTIONAL

Study Design: SINGLE_GROUP

Arm && Interventions

Group	Intervention	Description
HMD	No HMD used	We have several versions (listed below) of a headworn smartHMD. Each can provide verbal and/or tactile feedback to the user. Feedback is controlled by either the experimenter or by computer vision algorithms. 1. The ODG Smartglasses is commercially available. This system uses computer vision to guide a user to the destination using audio and/ or vibration feedback. 2) Tactile stimulator array. This device uses an Arduino Micro, HC-05 Bluetooth Module, L293D Motor Driver and coin vibration motors attached to a head-worn headband or glasses frame. The motors can be controlled directly by an experimenter or by computer vision algorithms. 3) Computer Vision Navigation prototyping system consists of two components: Intel RealSense camera and Alienware M15 laptop. All participants will receive the same 3 interventions: no HMD used, HMD worn but not active, and HMD worn and active. Participants may be tested with any or all of the systems described above.
HMD	HMD worn but not active	We have several versions (listed below) of a headworn smartHMD. Each can provide verbal and/or tactile feedback to the user. Feedback is controlled by either the experimenter or by computer vision algorithms. 1. The ODG Smartglasses is commercially available. This system uses computer vision to guide a user to the destination using audio and/ or vibration feedback. 2) Tactile stimulator array. This device uses an Arduino Micro, HC-05 Bluetooth Module, L293D Motor Driver and coin vibration motors attached to a head-worn headband or glasses frame. The motors can be controlled directly by an experimenter or by computer vision algorithms. 3) Computer Vision Navigation prototyping system consists of two components: Intel RealSense camera and Alienware M15 laptop. All participants will receive the same 3 interventions: no HMD used, HMD worn but not active, and HMD worn and active. Participants may be tested with any or all of the systems described above.
HMD	HMD worn and active	We have several versions (listed below) of a headworn smartHMD. Each can provide verbal and/or tactile feedback to the user. Feedback is controlled by either the experimenter or by computer vision algorithms. 1. The ODG Smartglasses is commercially available. This system uses computer vision to guide a user to the destination using audio and/ or vibration feedback. 2) Tactile stimulator array. This device uses an Arduino Micro, HC-05 Bluetooth Module, L293D Motor Driver and coin vibration motors attached to a head-worn headband or glasses frame. The motors can be controlled directly by an experimenter or by computer vision algorithms. 3) Computer Vision Navigation prototyping system consists of two components: Intel RealSense camera and Alienware M15 laptop. All participants will receive the same 3 interventions: no HMD used, HMD worn but not active, and HMD worn and active. Participants may be tested with any or all of the systems described above.

Primary Outcome Measures

Name	Time	Method
Time-to-Complete	2 hours	Duration from start of trial to subject completing their assigned task (seconds).

Secondary Outcome Measures

Name	Time	Method
Percentage of Preferred Walking Speed	2 hours	Percentage of preferred walking speed is a measure of how fast the participant is walking relative to their natural (or preferred) walking speed. Measure preferred walking speed was measured by having them walk with a researcher in an open area. Percentage preferred walking speed was obtained by dividing the measured walking speed during a trial by the preferred walking speed.
Number of Unintended Contacts With Obstacles and Walls	2 hours	This outcome measure only applies to the RGBD \& Jetson Indoor Test, as there were no walls/obstacles relevant in the other tests.
Number of Incorrect Turns	2 hours	Number of incorrect turns or responses to device cues during testing.
Number of Interventions	2 hours	Upon real-world application of the study protocol, the outcome measure originally conceived of as "requests for assistance" became pragmatically inseparable from "number of interventions" by study staff. Therefore, the two measures have been combined and are presented here as "number of interventions". The types of assistance requested could not be categorized, as they were not collected.