Monocular-Camera-Based Mobile Exergame for Type 2 Diabetes Prevention

Summary

Exergames have demonstrated potential as effective interventions for promoting physical activity and preventing type 2 diabetes (T2D), particularly among older adults. Kinect-based exergames, in particular, have been associated with improved adherence to exercise regimens and positive health outcomes. However, widespread implementation is limited by the high cost and reduced accessibility of the required hardware, restricting their use in home-based settings. Recent advances in computer vision have enabled the development of exergames using monocular camera systems, which may represent a cost-effective and scalable alternative. This study investigates the feasibility of monocular-camera-based exergames as a cost-effective and convenient alternative for home-dwelling individuals.

A total of 45 community-dwelling older adults aged 60-74 years, identified as high risk for T2D were recruited through local community health centers. Participants were randomly assigned to one of three groups (n = 15 per group): (1) Control group (traditional offline exercise with printed instructions), (2) Kinect group (Kinect-based exergames targeting aerobic capacity, balance, and strength), and (3) Monocular group (monocular-camera-based exergames using real-time 2D pose estimation). The intervention lasted six weeks, with participants completing three 30-minute sessions per week at home. Primary outcomes included exercise performance (completion rate and movement accuracy) and intrinsic motivation. Secondary outcomes included perceived enjoyment, challenge, and usability. Data were analyzed using one-way ANOVA.

Conditions

• Behavior Change Interventions

Interventions

BEHAVIORAL

Control group

Control Group consists of six motion videos, including: crossover steps, high knees, lateral raises, punching movements, downward leg punches from a standing position, and elbow-to-chest expansions

BEHAVIORAL

Kinect group

Kinect Group features the same software functionality and design as the application based on bullet-screen cameras, but the hardware is built using a Kinect sensor. The gamified platform is primarily developed and presented using Unity software. The Kinect 3D motion-sensing camera incorporates real-time dynamic capture and image recognition capabilities, offering new possibilities for interactive approaches to motion therapy

BEHAVIORAL

Monocular group

The game employs a virtual avatar to replicate users' movements, thereby fostering an immersive and engaging experience. A monocular camera captures users' movements in real time, which are analyzed through pose estimation algorithms and subsequently mapped onto the virtual avatar. Users interact with the game by following on-screen visual demonstrations, presented as either static images or animations, to perform the prescribed exercises. Movement accuracy is evaluated by the system, with scores awarded based on performance. To enhance user motivation and adherence, the game incorporates a reward system, where points earned through accurate execution can be redeemed for in-game rewards, such as unlocking background music, avatar customization options, and new virtual environments.

Sponsors & Collaborators

• Shanghai Jiao Tong University School of Medicine

  lead OTHER

Study Design

Allocation

RANDOMIZED

Purpose

PREVENTION

Masking

DOUBLE

Model

PARALLEL

Eligibility

Min Age

60 Years

Max Age

74 Years

Sex

ALL

Healthy Volunteers

Yes

Timeline & Regulatory

Start

2024-01-01

Primary Completion

2024-06-30

Completion

2024-08-01

Countries

• China

Study Locations