Page 82 - AIH-1-4

Artificial Intelligence in Health Robotics and Vivaldi AI for ALS assessment

et al. outlined how such robotic systems can be useful Milan. Participants met the diagnostic criteria for probable
19
for the management infectious diseases, including disease or definite ALS as outlined in the revised El Escorial
prevention and monitoring, clinical care, laboratory criteria. Inclusion criteria required ALS patients to
28
automation, and logistics, during the COVID-19 exhibit a minimum ALSFRS-R score of ≥2 points in both
pandemic. However, challenges such as difficult operation, speech and handwriting questions, with a total ALSFRS-R
maintenance, reliability, and high costs hinder their score of ≥18 points. In addition, all participants were
widespread adoption in private homes. capable of verbal communication to answer interview
Despite these challenges, numerous research efforts have questions. Patients displaying severe cognitive and
successfully tested assistive robotics in real-world settings, behavioral impairment, as determined by the Edinburgh
demonstrating their acceptance and usability in personalized Cognitive Assessment Scale at enrollment, were excluded
daily assistive plans. This progress highlights the potential from the study. Before participation, all patients provided
20
for these technologies to become integral components of informed consent, which was approved by the Local Ethics
patient care, improving the overall management and quality Committee, Grande Ospedale Metropolitano Niguarda,
of life for individuals with neurological disorders. 21 Milano, Italy (Protocol Number: 404-092019).
For instance, Ghafurian et al. examined the use of 4.2. Vivaldi AI system
22
social robots for dementia care, highlighting various Sanbot Elf is enhanced by the cutting-edge Vivaldi AI
levels of robot autonomy and assistive objectives such system, which equips it with advanced capabilities such as
as connecting patients with distant family members, speech recognition, facial expression analysis, and gesture
providing companionship, promoting health, and aiding understanding. Unlike conventional setups reliant on
in daily tasks. The authors underscored the need for remote cloud servers, this AI engine processes data locally
robots to be easy to use and maintain, and reliable, as on the robot itself, facilitating faster response times and
these factors significantly impact their adoption in private potentially reducing operational costs by minimizing data
homes. Saunders et al. reported the use of the Care- transfers. With Vivaldi AI at its core, Sanbot Elf adeptly
23
O-Bot 3 for complex interactions with elderly people, comprehends human communication, enabling it to
facilitating independent living and daily tasks. Schroeter engage with users in a natural and intelligent manner. Key
et al. reported the usability of the social assistive robot, functionalities include: (1) speech recognition, facilitating
24
which provides questionnaires and cognitive stimulation effortless voice interaction with users; (2) natural language
games to monitor cognitive decline to be used in a home processing, enabling the robot to understand and respond
setting. Instead, Fischinger et al. developed a care robot to human language nuances; (3) facial expression
25
(namely Hobbit robot prototype) for aging by means of analysis, allowing Sanbot Elf to discern user sentiment
fall prevention/detection. The Hobbit project combined by interpreting facial cues, thereby tailoring its responses
research from robotics, gerontology, and human-robot accordingly; (4) gesture understanding, enhancing the
interaction to develop a care robot, which is capable of fall fluidity of communication between the robot and users;
prevention and detection as well as emergency detection and (5) seamless communication, ensuring smooth and
and handling. Moreover, to enable daily interaction with intuitive interaction between Sanbot Elf and its human
the robot, other functions were added, such as bringing counterparts.
objects, offering reminders, and entertaining. Casey
et al. have demonstrated that robots could promote social 4.3. Robotic administration of ALSFRS-R
26
connection and reduce loneliness among dementia patients questionnaire
in a long-term residential setting. Finally, Neerincx et al.
27
designed a personalized assistive robot for patients with For this study, the Sanbot Elf unit is equipped to administer
cognitive impairments, taking into account their daily the ALSFRS-R questionnaire using a structured format
needs and characteristics such as their personal cognitive, allowing patients to answer questions with either “yes” or
emotional, and psychological status as well as their “no.” This approach is intentionally adopted to minimize
cultural background, which are not always easily modeled ambiguity in patient responses. Patients have the option
according to well-defined classifications. to respond either verbally or by utilizing the touchscreen
display, particularly beneficial for those experiencing
4. Data and methods dysarthria or hypophonia. The Vivaldi AI system guides
patients through the ALSFRS-R questionnaire, dynamically
4.1. Participants adjusting the sequence of questions based on the patient’s
This study enrolled a cohort of 28 ALS patients who responses. This ensures a personalized experience tailored
underwent evaluation at the Nemo Clinical Center in to each individual’s needs. Throughout the questionnaire,

Volume 1 Issue 4 (2024) 76 doi: 10.36922/aih.3732