The StepKinnection project explores the use of interactive video games to reduce the risk of falling in elderly.
Our approach consists on developing games that combine the fun factor inherent in video games with a clinical assessment tool.
While the entertainment factor engages the person in to exercising, the incorporation of a clinical test allows for the continuous assessment of health outcomes.
Falls Prevention Game
A Kinect-driven stepping game for the elderly, based on stepping exercises that train specific cognitive and physical abilities associated with falls.
Automated Risk Assessment
This system combines a set of suitable age-related features and an embedded clinical test for fall risk assessment.
At the current stage of the project, the following two systems have been created.
A Game for Falls Prevention
A Kinect-driven stepping game for the elderly, based on stepping exercises that train specific cognitive and physical abilities associated with falls. This system combines a set of suitable age-related features and an embedded clinical test for fall risk assessment.
A Risk Assessment Tool
An interactive system for the elderly that incorporates mechanisms to simultaneously perform a hybrid clinical test for fall risk assessment. This clinical test includes a simple stepping task along with three voice-enabled cognitive activities allowing for the assessment of stepping performance under the dual-task paradigm.
Awards
This project received several scientific awards, which include:
HIC 2012
Branko Cesnik Award
Best student paper award for the paper 'Exergames for the Elderly: Towards an Embeded Kinect-based Clinical Test Validation'
HIC 2015
Branko Cesnik Award
Best student paper award for the paper 'StepKinnection: A fall prevention game mindfully designed for the Elderly'
IE 2018
Best Paper Award at Interactive Entertainment
The UTS Games Studio has won the Best Paper award at the Interactive Entertainment conference, which was run as part of the 2018 Australasian Computer Science Week held at the University of Queensland in sunny Brisbane, Australia.
Our Team
Dr. Jaime Garcia
Serious Games to Improve the Physical Health of the Elderly
Dr. Karla Felix Navarro
Mobile and Wireless Communication Technologies
A/Prof Yusuf Pisan
Game Design, Game Development, Graphics, iPhone Programming, Artificial Intelligence, C#
Dr. Chek Tien Tan
Artificial Intelligence in Games, Computer Game Design and Programming, Data Structures and Algorithms
StepKinnection is an interactive video game that delivers stepping exercises for the elderly. This system combines a set of appropriate age-related features and a hybrid version of the Choice Stepping Reaction Time (CSRT) task, a time-based clinical test that has shown to reliably predict falls in older adults.
The combination of these two dimensions make this game potentially useful as an effective tool to reduce the risk of falling in older people as:
it directly appeals to the aged cohort;
it trains specific physical and cognitive functions associated with falls; and
it allows for a continuous assessment of their health outcomes in order to evaluate their progression.
To the best of our knowledge, none of these characteristics is available in current commercial games.
Game Description
In this game, the player is an explorer who travels around the globe visiting colorful countries, hunting for treasures and seeking different adventures. Each country presents a challenge where the player gets to experience their traditional music and collect exotic fruits. Completing each challenge takes the player one step closer to winning a trophy. However, the further they travel the more difficult it gets. Game play starts with a series of basic levels where players will have the chance to familiarize with the game and coordinate their movements accordingly. Once they have finished these levels, players can move up to more challenging ones related to speed, precision and cognitive complexity.
In order to play the game, the player needs to stand in front of the TV facing the Kinect. Shortly after, the main menu will be presented (Figure 1), where players can select from a list of counties to visit. The player can wave either hand to move the cursor to navigate the menu (Figure 2). Once the level is selected the user moves to the main stepping task.
The Stepping Task
Throughout the game, fruits will appear on the screen every now and then, and the player is expected to collect them. In order to achieve this, the player needs to reach the fruits by stepping on them (Figure 3).
As the user moves through the levels, the speed of the appearance and the size of the fruits decrease. This is to encourage players to perform quicker and more accurate steps (i.e., increase in skill) as they advance to higher levels. This hence reinforces the ability to take proactive steps that could help an individual to regain balance and avoid falls. Also, as the stepping area decreases, the user needs to be more coordinated to be able to step on the fruits.
Training Motor Inhibition
In the mid-levels, a lady bug might randomly appear on the screen (Figure 4). Stepping on the lady bug will take 2 penalty points off their current score reducing their chances of winning. However, if the player remains in position, 1 point will be awarded.
This motor inhibition task is incorporated with the purpose of adding varied difficulty to the game by slightly increasing its cognitive demand. According to [8], adequate motor inhibition plays an important role in avoiding falls. Training this ability is therefore ideal for situations where avoiding an obstacle can prevent a fall from occurring.
Stimulating the Ability to Take Quick Reactive Steps
In the higher levels, dollar coins will randomly appear on the screen for split-second amounts of time (Figure 5). These coins are bonus points that can help players to move faster in the game, with the purpose of encouraging them to step faster. For each dollar coin that they collect, 2 bonus points will be added to their current score to reward the player. This feature trains the ability to respond quickly to a hazardous situation. Ideal for circumstances where the person has initiated a step but the environment suddenly changes
For all the above stepping tasks in the game, the accuracy of the responses is automatically processed by the hybrid clinical test for fall risk assessment that is embedded in the game.
The Step Kinnection Test for Fall Risk Assessment
Overview
The StepKinnection Test is an interactive system for the elderly that incorporates mechanisms to simultaneously perform a hybrid clinical test for fall risk assessment.
This clinical test includes a simple stepping task along with three voice-enabled cognitive activities allowing for the assessment of stepping performance under the dual-task paradigm.
Test Description
The Step Kinnection Test, a game-like system that delivers step training exercises to older adults and simultaneously measure stepping performance through the use of a hybrid version of the Choice Stepping Reaction Time (CSRT) task, a time-based clinical test that has shown to reliably predict falls in older adults.
The main motivation for choosing the Kinect as the primary input device is that it allows for:
a wider degree of freedom for the user;
an intuitive and natural interaction with the game as no controllers or wearable sensors are required for its operation;
a better provision of feedback allowing the display of a full body avatar to mirror the users’ movements.
All these features are ideal for elderly users with minimal or no computer literacy.
In addition to this, Step Kinnection also allows for the collection of spatial information such as postural sway and stepping accuracy. More importantly, the assessment of stepping performance under the dual task paradigm can also be achieved through the incorporation of a series of cognitive activities [10]. Poor dual tasking has been frequently associated with falls and balance impairments in older people, providing evidence for the importance of specific cognitive functions in postural stability [10]. These features make this system potentially useful in actual clinical practice to evaluate various dimensions involved in the diagnosis of fall risk in older people, all in a single system.
The Stepping Task
In order to start the game, the player is required to stand in front of a computer screen or TV connected to a Kinect PC. The representation of the player in the system is a pair of shoes mirroring the person’s feet. Six symmetrically distributed square-shaped panels are then drawn on the screen surrounding the player’s avatar(Figure 1). When one of the panels changes its color to green on the screen, the player is expected to step on it in space and back to the center as quickly as possible. As soon as the player returns to the initial position the process starts over. The sequence is presented randomly as well as the time between trials so that the user is unable to anticipate the time and location of the next stimulus. It is worth mentioning that these virtual panels are dynamically located based on the user’s height, making the stepping task equally challenging for short and tall participants. While playing, time based measurements such as reaction times are simultaneously collected. Also the positioning of the foot is recorded in order to determine the accuracy on stepping as well as the step length.
The Cognitive Tasks
In addition to the stepping game, three voice-controlled concurrent cognitive activities were incorporated to assess the performance of the patient under differing cognitive and motor conditions concurrently. The increased cognitive load affects the user performance while stepping, resulting in noticeably slower reaction times for users that are likely to fall.
The three activities are:
Read the Word :
During this task the user is required to say the color out loud while performing the stepping exercises. As the color of the word and its semantic meaning are identical, this task creates a minimally increased cognitive load for the user.
Name the Colour:
For this task the semantic meaning of the word and the color of the word do not match. Once again the user has to say the color out loud, but in this case there is interference between the meaning and the color of the word. While the mind automatically determines the meaning of the word, the player actually needs to identify the color that the word is written in. This means the player needs to consciously re-evaluate his/her instinctive response. This interference, also known as the Stroop interference, results in a delay and the extra processing required normally results in a slowing down of the stepping test performance.
Maths Workout
This task requires the user to answer a math question that is read by the system. While there is no interference effect as with the Stroop test, the user is still required to interpret what they have heard to answer the question.
For all the above tasks, the accuracy of the answer is automatically processed by the voice recognition system built into the Kinect.