Intelligent machines and mental health in the era of COVID-19.

The idea of a network of small devices that would be able to connect each other, appeared in the early 80s. In a prophetic article, Mark Weiser,1 described such a connection, that it is now known under the term of Internet of Things (IoT). In a broadest sense, the term IoT encompasses everything connected to the internet, but it is increasingly being used to define objects that "talk" to each other, creating a network from simple sensors to smartphones and wearables connected. During the recent years this network of communicating devices has been combined with other technological achievements, and particularly with the Virtual Reality (VR)2 and the Artificial Intelligence (AI).3 The emerge of COVID-19 pandemic in 2019, resulted to the poor response and healthcare failures of many countries globally.4 One of the main reasons for such a failure, was the inability of accurate data collection from different sources. Apparently, it was the first time, humanity realized the need for massive amounts of heterogeneous data to be collected, interpreted, and shared. Amid the ongoing COVID-19 pandemic, several innovators and public authorities are looking to leverage IoT tools to reduce the burden on the healthcare systems.5 Mental health is one of the areas that seems to benefit the most of such technologies. A significant decrease of the total amount of ER visits and a dramatic increase of internet access from the patients and care givers along to the development of applications for mental health issues, followed the outbreak of SARS-CoV-2.6 Such technologies proved to be efficient to help mentally ill patients and pioneer the path in the future. Probably the most obvious use of these emerged technologies is the improvement of the telehealth options. Patients who suffer from mental illness face significant problems towards the continuity of care during the crisis.7 Nonetheless, they usually have other health problems, that deprive them from an equitable health care provision. Improved telehealth platforms can give them a single point access to address all their problems. The use of electronic health records can reduce the fragmentary health services and improve the outcome.8 However, this is only the beginning. The COVID-19 crisis and the subsequent social isolation, to reduce both the contamination and the spread of the disease, highlighted the necessity for providing accurate and secure diagnoses and treatments from a safe distance. Virtual reality combined with IoT and AI technologies seem to be a reliable alternative to the classic physical and mental examination and treatment in many areas of mental and neurological diseases.2 These novel techniques can spot the early signs and detect mental illnesses with high accuracy. However, caution and more work are required to bridge the space between these recently thrived technologies and mental health care.7 It is worth mentioning, that internet-oriented health care procedures can also help to reduce the gaps caused by the stigma of mental illness. For example, the development of AI chatbots (an application used to chat directly with a human) can alleviate the fears of judgment of the help seeking persons and provide the professionals with a supplemental support toward improved services to their patients.9 A final remark for conclusion. Humanity is more and more depended to the "intelligent" machines. However, we must not forget that we humans are responsible to set the rules of such co-existence.

Personalized Virtual Reality Human-Computer Interaction for Psychiatric and Neurological Illnesses: A Dynamically Adaptive Virtual Reality Environment That Changes According to Real-Time Feedback From Electrophysiological Signal Responses.

Virtual reality (VR) constitutes an alternative, effective, and increasingly utilized treatment option for people suffering from psychiatric and neurological illnesses. However, the currently available VR simulations provide a predetermined simulative framework that does not take into account the unique personality traits of each individual; this could result in inaccurate, extreme, or unpredictable responses driven by patients who may be overly exposed and in an abrupt manner to the predetermined stimuli, or result in indifferent, almost non-existing, reactions when the stimuli do not affect the patients adequately and thus stronger stimuli are recommended. In this study, we present a VR system that can recognize the individual differences and readjust the VR scenarios during the simulation according to the treatment aims. To investigate and present this dynamically adaptive VR system we employ an Anxiety Disorder condition as a case study, namely arachnophobia. This system consists of distinct anxiety states, aiming to dynamically modify the VR environment in such a way that it can keep the individual within a controlled, and appropriate for the therapy needs, anxiety state, which will be called "desired states" for the study. This happens by adjusting the VR stimulus, in real-time, according to the electrophysiological responses of each individual. These electrophysiological responses are collected by an external electrodermal activity biosensor that serves as a tracker of physiological changes. Thirty-six diagnosed arachnophobic individuals participated in a one-session trial. Participants were divided into two groups, the Experimental Group which was exposed to the proposed real-time adaptive virtual simulation, and the Control Group which was exposed to a pre-recorded static virtual simulation as proposed in the literature. These results demonstrate the proposed system's ability to continuously construct an updated and adapted virtual environment that keeps the users within the appropriately chosen state (higher or lower intensity) for approximately twice the time compared to the pre-recorded static virtual simulation. Thus, such a system can increase the efficiency of VR stimulations for the treatment of central nervous system dysfunctions, as it provides numerically more controlled sessions without unexpected variations.

Comparison between Full Body Motion Recognition Camera Interaction and Hand Controllers Interaction used in Virtual Reality Exposure Therapy for Acrophobia.

Virtual Reality has already been proven as a useful supplementary treatment tool for anxiety disorders. However, no specific technological importance has been given so far on how to apply Virtual Reality with a way that properly stimulates the phobic stimulus and provide the necessary means for lifelike experience. Thanks to technological advancements, there is now a variety of hardware that can help enhance stronger emotions generated by Virtual Reality systems. This study aims to evaluate the feeling of presence during different hardware setups of Virtual Reality Exposure Therapy, and, particularly how the user's interaction with those setups can affects their sense of presence during the virtual simulation. An acrophobic virtual scenario is used as a case study by 20 phobic individuals and the Witmer-Singer presence questionnaire was used for presence evaluation by the users of the system. Statistical analysis on their answers revealed that the proposed full body Motion Recognition Cameras system generates a better feeling of presence compared to the Hand Controllers system. This is thanks to the Motion Recognition Cameras, which track and allow display of the user's entire body within the virtual environment. Thus, the users are enabled to interact and confront the anxiety-provoking stimulus as in real world. Further studies are recommended, in which the proposed system could be used in Virtual Reality Exposure Therapy trials with acrophobic patients and other anxiety disorders as well, since the proposed system can provide natural interaction in various simulated environments.

Emotional stimulation during motor exercise: An integration to the holistic rehabilitation framework.

Over the recent years, plenty of studies have been carried out concerning disabilities and rehabilitation. However, very few of them have focused on proposing or creating a holistic framework meant to maximize the merits of rehabilitation treatments. More importantly, an even smaller number has taken into consideration the immense impact of emotions and the crucial role they display concerning patients' performance during rehabilitation. The purpose of this study is to examine and highlight the direct connection of patients' performance quality with an artificially induced positive or negative emotional state. For that reason, we conducted an experiment regarding the convergence of emotions with the motor exercise. Specifically, our study was conducted with 45 participants and the GAPED database was employed as an emotional classifier. The effect of the visual stimulus was combined with a simple bicep exercise which introduces the motor element in our study. Statistical analysis of the yielding EMG & Acellerometer signals demonstrated a considerable difference in the quality of the physical activity of each subject induced by different emotional triggers. As a result, it is pivotal to integrate a holistic approach to rehabilitation in future clinical trials, aiming to reinforce the patient involvement and engagement process and thus the efficacy of the treatment.