Tech for Aging and Dementia Session for International Members on Nov 23

Welcome & Introduction
Presenters:
Pattie Maes, Prof, Fluid Interfaces
Hae Won Park, PhD, Personal Robots

Description:
The elderly population is growing and with it, the needs for better technologies to detect, monitor and assist people with Dementia. This event will survey some of the emerging technologies developed at the Media Lab which may offer novel solutions, including: social robots, machine learning, and wearable sensor technologies.

Project: Rhizome
Presenter: Mostafa ‘Neo’ Mohsenvand (PhD Candidate), Fluid Interfaces
Description: Rhizome is a digital reminiscence therapy and monitoring solution designed to evaluate long-term loss of explicit memories, and to support patients with memory decline. Rhizome resembles a digital photo-album with memory testing functionalities that allow the patient and their family and friends to import and annotate memories. Rhizome is being built and tested with the help of NTT Data.

Project: Older Adult Co-Design
Presenter: Anastasia K. Ostrowski, Personal Robots
Description: The Personal Robots group engages with older adults, above the age of 70, to design social robots. Older adults collaborate with researchers through interviews, art-making, robot rapid prototyping, design guideline generation, and living with robots resulting in qualitative and quantitative data informing future design of social robots. We discuss several areas where robots could assist or facilitate, including emotional wellness, social connection, and memory.

Project: Emotional Wellness
Presenter: Sooyeon Jeong, Personal Robots
Description: With global rates of older adults expected to increase in the next 10-30 years, it is critical to meet their demand for physical and mental health. We present a socially  embodied AI companion which delivers twelve positive psychology sessions in older adults’ homes over a four week period. Based on the results from a previous study with college students, in which participants' personality traits had a significant impact on the outcomes, we chose to investigate the impact of two different robot personas (coach-like vs. companion-like) on the intervention efficacy. Our robotic intervention will aim to provide companionship, as well as helpful wellbeing interventions, to older adults from both independent-living and assisted-living facilities.  

Project: Medication Adherence
Presenter: Sharifa Alghowinem, Personal Robots
Description: Statistics published by the World Health Organization show that nonadherence to medication caused 50% of the treatment failure,125,000 premature deaths, and up to 25% of hospitalizations.  We introduce a social agent technology that not only reminds the user about their medication schedule, but also provides personalized support as a motivating ally for the purpose of medication adherence. The agent (Jibo) monitors the pill bottle usage using RFID technology, in order to use this information to support proper medication intakes. In this first step, the two systems (Jibo and RFID) are integrated and Jibo interacts with the user given the bottle tracking results and the medication knowledge is implemented. The initial implementation shows the feasibility of the proposed system, where we believe it will be effective in medication adherence. 

Project: Brain Switch
Presenter: Nataliya Kosmyna, PhD, Fluid Interfaces
Description: The Brain Switch is a closed-loop brain-computer system allowing for real-time correspondence of simple user needs to a caretaker, non-verbally. The Brain Switch is a lightweight, wearable, and wireless system which aims to help restore communication to those with physical challenges (ALS, CP, SCI). Brain Switch is a part  of the AttentivU system. Its architecture consists of a wearable, wireless electroencephalography system, which is comfortable to the user and can be worn over extended periods of time (no sticky electrodes, 8 hours+ of active use, 2 hours of charging time). It supports the needs of users with limited mobility, such as those who require or wish to remain in bed, have limited neck muscle support, etc.  

In addition, two mobile applications are being provided to the families; One being a notification app used by the caretaker for remote insights (e.g., what the user needs now), and the othe used by or near the patient. These applications will connect to the devices through a server. The patient's app streams packets of raw electroencephalography data from a  brain computer interface back to the server. With this information, a convolutional neural network is trained and used to classify mental states like imagery, as well as being able to follow the attention of the user in real-time. This information can then be forwarded to the waiting applications. Enabled by the cloud platform, this system is able to work anywhere and on a multitude of devices, computationally unlimited. 

Project: DementAI
Presenter: Utkarsh Sarawgi, Fluid Interfaces
Description: According to the World Health Organization, Alzheimer's disease is estimated to affect around 50 million people worldwide (rising rapidly). The WHO also states that Alzheimer’s currently produces a global economic burden of nearly a trillion dollars, with 63% of people affected by dementia living in low- and middle-income countries. This calls for scalable, cost-effective, and robust methods for detection of Alzheimer's Dementia (AD). DementAI is an open-source platform for modeling risk stratification of Alzheimer's Dementia using spontaneous speech through a mobile service, powered by privacy-protected and interpretable AI.