Using a Personalised Socially Assistive Robot for Cardiac Rehabilitation: A Long-Term Case Study

Abstract

This paper presents a longitudinal case study of Robot Assisted Therapy for cardiac rehabilitation. The patient, who is a 60-year old male that suffered a myocardial infarction and received angioplasty surgery, successfully recovered after 35 sessions of rehabilitation with a social robot, lasting 18 weeks. The sessions took place directly at the clinic and relied on an exercise regime which was designed by the clinicians and delivered with the support of a social robot and a sensor suite. The robot monitored the patient's progress, and provided personalised encouragement and feedback. We discuss the recovery of the patient and illustrate how the use of a social robot, its sensory systems and its personalised interaction was instrumental to maintain engagement with the programme and to the patient's recovery. Of note is a critical event that was promptly detected by the robot, which allowed fast intervention measures to be taken by the medical staff for the referral of the patient for further surgery.

A Socially Assistive Robot for Long-Term Cardiac Rehabilitation in the Real World

“…Most of these results are due to sessions where this alert was triggered for a high number of times due to malfunctioning of the robot. While this limitation may have caused false positives, this alarm is very important and vital to the clinicians for immediately detecting an emergency (e.g., symptoms of dizziness or abnormal blood pressure), such that the patient could be supported in a rapid manner (Irfan et al, 2020 ).…”

“…Continuous monitoring of the heart rate enabled providing immediate feedback to alert the medical staff in case of emergencies, which enabled rapid intervention (e.g., through decreasing the physical activity intensity parameters) from them. Hence, despite the fact that the robot malfunctioned and gave an excessive number of alerts on some occasions, this alert proved to be valuable to the medical staff (Irfan et al, 2020 ).…”

“…A second limitation is that our robot's behaviors were identical for each participant, which led some participants to report that they would like the robot to be more personalized. We are currently exploring personalization strategies (Irfan et al, 2020 ) (e.g., recognizing patients, using their name, and referring to progress in the previous sessions) to address these concerns and improve the adaptation to the robot without the perception of repetitiveness or boredom. Moreover, the patients in the Control condition did not receive motivational or physiological verbal feedback to emulate the conventional CR sessions.…”

“…We aimed to investigate the effects of using a socially assistive robot that uses these sensory values to give verbal feedback to the patient, in comparison to a Control condition, where the sensory values are only displayed on a tablet-based Graphical User Interface (GUI), i.e., no verbal feedback is given to the patient. Initial user studies (Lara et al, 2017 ; Casas et al, 2018a , b ; Casas et al, 2018c ; Casas et al, 2020 ; Aguirre et al, 2020 ; Irfan et al, 2020 ) validated the sensor interface and the robot within the laboratory and clinical settings based on case studies. In Casas et al ( 2019 ), we compared the perceptions of the patients that have conducted the cardiac rehabilitation programme with the robot to the expectations of the patients that have not interacted with the robot through a control group.…”

Multi-modal Open World User Identification

“…the mathematical formulations and assumptions behind the models that were not addressed in Reference [43]. In addition, we present our findings from applying the optimised models in long-term HRI experiments in the real world [41][42][43]. Correspondingly, we make the following contributions (source code, multi-modal dataset, trained models, and results on the dataset are available online 2 ):…”