IEEE Robotics & Automation Magazine - March 2023 - 68

isometric arm lift. With this aim, as shown in Figure 5(b),
the user first moves the mouse pointer (via arm motion) to
the position indicated on the screen (one of the orange-colored
circles) and then actively lifts the arm against the
device weight as much as possible while maintaining the
position. The green-colored circle size indicates the variation
of the measured lifting force.
■ The control of the vertical force (COF) assessment for the
arm game [Figure 5(c)] quantitatively assesses the capacity
of the person to lift the arm but in a relaxed position (no
movement on the table is needed).
■ The assessment of hand function (AHF) consists of two
subassessments: prono-supination movement of the wrist
and hand opening/closing [Figure 5(d)]. The small rectangle
in the gray arc moves as the user performs the pronosupination
movement with respect to the reference
position (the center of the arc), and the reachable range of
prono-supination (ROPS) is marked with a green-colored
arc. The gray circle in Figure 5(d) is used to monitor the
range of finger opening and closing force (ROFF). The
pressure resulting from FSR sensors in the hand module
changes the size of the circle to represent the hand grasp
and release function. The final diameter and arc length of
the green-colored fields are registered as the wrist and
hand motion capacity.
In the trials previously performed [16], [17], it has been
observed that during the assessment, users make a special
effort to reach their maximum capability. The use of maximum
values can be demotivating in training games because
users cannot achieve those values consistently during the rehabilitation
sessions. Hence, 70% of the maximum measured
values from the ROM, ROF, and ROPS assessments applies
to the training games, while 50% of the maximum value from
the ROFF assessment achieves those values consistently during
the rehabilitation sessions. Hence, 70% of the maximum
measured values from the ROM, ROF, and ROPS assessments
applies to the training games, while 50% of the maximum
value from the ROFF assessment is used at the request of the
physiotherapist.
TRAINING GAMES
As stated previously, 10 serious games have been developed
for training. The games use the information transmitted from
the robotic module to determine the position of a cursor on
the screen, which is relative to the calculated ROM in AAA.
Each game targets different aspects of function, so the platforms
will offer only the games available for the movements
to be trained (Table 3). The features of each game are
explained next.
■ Discover the Picture: Discover the Picture [Figure 6(a)]
presents an obstructed picture that is uncovered by the
user during planar arm reach movements in each of eight
directions.
■ Shopping Cart: In Shopping Cart [Figure 6(b)], the goal of
the game is to pick up objects that appear on the screen.
When the objects are picked up, money is spent.
■ Shapes: In Shapes [Figure 6(c)], the user is asked to move
objects located on the lower portion of the screen and pair
them with their respective match in the upper portion.
■ Memory: In the Memory game, the user is asked to find
pairs of identical cards.
■ Puzzle: In Puzzle [Figure 6(d)), pieces must be assembled
to complete the image.
■ Words: In Words [Figure 6(e)], some letters of a word
appearing are missing, and the user is asked to complete
the word.
■ At the Bar: In At the Bar [Figure 6(f)], the user has to take
the bottle and fill in the glass through the coordination of
both grasping and prono-supination movements.
■ Solitaire: In Solitaire, the user has to complete the solitaire
by placing the cards in the correct place. This game is recommended
for users with relatively high dexterity and
long concentration. At least 5-10 min are required.
(a)
(b)
■ Online Games: Existing online games can be played using
the robotic system as a mouse.
Each training game, except for the online games, has five levels
that are automatically updated based on the evolution of the
user or updated manually by the therapist. The higher the level
is, the higher involvement of motor and/or cognitive function
is needed. The automatic update of the level is done depending
on the performance, which is defined by the percentage of the
obtained score over the maximum possible. The score calculation
for each game is shown in the last column of Table 3. The
level is automatically updated in the event that the user obtains
a performance higher than 80% twice or 100% once.
(c)
(d)
FIGURE 5. AAA games. (a) A range of motion (ROM) assessment.
(b) A range of vertical force (ROF) assessment. (c) A control of the
vertical force (COF) assessment. (d) Assessment of hand function
(AHF).
68 IEEE ROBOTICS & AUTOMATION MAGAZINE MARCH 2023
USABILITY EVALUATION
AA has been designed to facilitate robotic-assisted rehabilitation
at home, allowing the user to continue effective therapy
with low supervision while maintaining safety. AA was evaluated
in multiple clinical settings; results show that it is well
accepted by therapists and users and makes the therapy more
enjoyable and motivating [17], [18].
IEEE Robotics & Automation Magazine - March 2023

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2023
