IEEE Robotics & Automation Magazine - March 2023 - 91

According to the normal functioning
model of language, both tools identify the
patient's language/speech deficits.
INTERVENTION
EG
The EG was composed of patients who
conducted speech therapy with the
SmartCARE system; they were free to
use the system as much as they wanted
with no limitations (at least three a week,
five weeks), and the session lasting was
linked to the number of exercises prescribed.
A trained technician was
responsible for installing all of the equipment.
The main PC was wired, taking
care to hide all cables behind the TV
screen as much as possible to ensure a
neat and tidy appearance of the system.
Figure 5 shows the geographic distribution
of SmartCARE users of the EG.
In addition, all subjects were trained to
"
BOTH INTENTIONS TO
USE AND PERCEIVED
USEFULNESS APPEAR
TO BE POSITIVELY INFLUENCED
BY AESTHETICS
AND CONTROL FACTORS,
SUGGESTING
HOW THE GRAPHICAL
USER INTERFACE AND
USER EXPERIENCE PLAY
A CRITICAL ROLE IN THE
CHOICE OF A HEALTHCARE
SYSTEM.
„
use the system and all exercises with the help of a professional
team of experts. Patients were able to interact with an adaptive
interaction mode (Figure 6) based on gesture recognition with
the mouse.
The proposed system was used to execute speech-therapy
exercises and improve cognitive functions with the serious
games described in a previous article [14].
CG
The CG was composed of patients who
underwent traditional speech therapy
(three a week, five weeks, 45 minutes per
session). The treatments were composed
of exercises that addressed the same areas
of comprehension as in the EG (listening
comprehension of words, especially
nouns, verbs, and syllables; listening comprehension
of sentences; reading comprehension
of words, nouns, and verbs;
comprehension of written sentences) with
the modality of face-to-face presence.
STATISTICAL ANALYSIS
Statistical descriptive analysis was conducted
to assess the usability and acceptance
of the platform during the
experiments. For the assessment of technology
acceptance, Cronbach's alpha and
Kendall's tau-b correlation coefficients
were calculated to assess reliability, discriminant
validity, and correlations
between responses of all pairs of domains from TAM.
For clinical data, statistical analyses were performed
using IBM SPSS Statistics 25 (IBM, Armonk, NY, USA).
Wilcoxon signed-rank tests were used to examine the differences
between pre- and postrehabilitation outcomes
of the two groups (EG and CG). Statistical analysis was
performed with a 95% confidence interval, and a p value
< 0.05 (two-sided) was considered statistically significant.
RESULTS
USABILITY AND ACCEPTANCE OF THE PLATFORM
The analysis of usability and acceptance of the SmartCARE
system was based on data from the EG (dropout data were
not included).
TABLE 3. Summary of the SmartCARE usage.
USAGE
Number of days since the start of the pilot
phase (days)
Total number of performed sessions at home (n)
Total time to execute the exercises (interactive
movement and measurement) (hours)
Average time to execute all of the prescribed
exercises per each session (minutes)
Total exercises executed by all patients during
the pilot period (n)
FIGURE 6. System@Home during touchless mode execution of a
" visual comprehension of sentences " exercise.
Average elapsed time spent by a patient to complete
a therapy session, including reading the
instructions, showing demonstration videos, and
execution of the exercises (hours)
RESULT
139
312
133.02
25.58
3,579
37.05
MARCH 2023 IEEE ROBOTICS & AUTOMATION MAGAZINE
91
IEEE Robotics & Automation Magazine - March 2023

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