Description
Cognitive Behavioral Assistive Technology Team (https://aip.riken.jp/labs/goalorient_tech/cogn_behav_assist_tech/) at RIKEN AIP
Speaker 1: Mihoko Otake-Matsuura (20 min)
Title: Overview of the Cognitive Behavioral Assistive Technology (CB-AT) Team
Abstract: AI nurturing or complementing intelligence of human are becoming much more important than ever before. Cognitive Behavioral Assistive Technology (CB-AT) Team puts emphasis on developing CB-AT which promotes cognitive reserve for prevention of cognitive decline and dementia of older adults which damage the intelligence of human necessary for social life. Research topics include: 1) Cognitive behavioral assistive systems for interactive communication of older adults (Seiki Tokunaga), 2) Analysis and modelling technology of conversational, physiological and psychological data (Takuya Sekiguchi, Masato S. Abe, Tomasz M. Rutkowski, Hikaru Sugimoto), 3) Clinical study of the developed systems evaluating whose effects on human (Otake-Matsuura, M. et al, 2021).
Reference:
Otake-Matsuura M, Tokunaga S, Watanabe K, Abe SM, Sekiguchi T, Sugimoto H, KIshimoto T, Kudo T (2021) Cognitive Intervention through Photo-Integrated Conversation Moderated by Robots (PICMOR) Program: A Randomized Controlled Trial. Frontiers in Robotics and AI, 8:633076. https://doi.org/10.3389/frobt.2021.633076
Speaker 2: Seiki Tokunaga (20 min)
Title: Dialogue-Based System with Photo and Storytelling for Older Adults: Toward Daily Cognitive Training at Home
Abstract: In a super-aged society, dementia is a severe problem in which older adults suffer from symptoms to maintain daily life.A wide variety of systems have been developed for older adults; however, to our knowledge, few of them encourage communication and aim to train cognitive functions at home. Additionally, home-based experimentation has become crucial in COVID-19 situations to keep participants safe and provide more chances for older adults living alone to communicate.In this presentation, we report on how to develop home-based experiments using a novel dialogue system (Tokunaga S, Tamura K, Otake-Matsuura M, 2021). We also present how to conduct a home-based experiment and its current result.
Reference:
Tokunaga S, Tamura K, Otake-Matsuura M (2021) A Dialogue-Based System with Photo and Storytelling for Older Adults: Toward Daily Cognitive Training. Frontiers in Robotics and AI, 8: 644964. https://doi.org/10.3389/frobt.2021.644964
Speaker 3: Takuya Sekiguchi (20 min)
Title: Types of social networks and starting leisure activities in later life
Abstract: Health benefits of older adults’ leisure activities have been reported. Literature has also shown that starting activities even in later life is beneficial, which indicates that it is reasonable to encourage older adults to be active. Social networks have various functions to facilitate older adults to be active. The present study extracted the patterns of social networks among older adults who were not engaged in leasure activities, and examined which patterns increase the likelihood of their starting activities three years later (Sekiguchi T, Kondo K, Otake-Matsuura M, 2021).
Reference:
Sekiguchi T, Kondo K, Otake-Matsuura M (2021) Types of social networks and starting leisure activities in later life: A longitudinal Japan Gerontological Evaluation Study (JAGES). PLOS ONE 16(7): e0254828. https://doi.org/10.1371/journal.pone.0254828
Speaker 4: Hikaru Sugimoto (20 min)
Title: Neural mechanisms of how cognitive functions are modulated in social contexts
Abstract: Humans are social beings. In daily life that usually requires social interactions with others, their minds, including cognition and affect, would be influenced by various forms of social activities. However, there is still limited evidence regarding neural mechanisms of how cognitive functions are modulated in social contexts. Here, I will report our recent publications on neuroimaging studies that investigated this issue. In this talk, I will show fMRI evidence regarding the neural mechanisms of how episodic memories are influenced by the receipt or anticipation of rewards derived from social interactions with others, such as victories in competitions [1, 2]. In addition, I will report our preliminary findings from a resting-state fMRI study that aimed to provide candidate neural mechanisms underlying the beneficial effect of a group conversation-based intervention program on verbal fluency in older adults [3].
References:
[1] Sugimoto H, Dolcos F, Tsukiura T. Memory of my victory and your defeat: Contributions of reward- and memory-related regions to the encoding of winning events in competitions with others, Neuropsychologia 152, 207733 (2021). https://doi.org/10.1016/j.neuropsychologia.2020.107733.
[2] Sugimoto H, Shigemune Y, Tsukiura T. Competing against a familiar friend: Interactive mechanism of the temporo-parietal junction with the reward-related regions during episodic encoding, Neuroimage, 130 261-272 (2016). https://doi.org/10.1016/j.neuroimage.2016.02.020
[3] Sugimoto H, Kawagoe T & Otake-Matsuura M. Characteristics of resting-state functional connectivity in older adults after the PICMOR intervention program: a preliminary report. BMC Geriatr 20, 486 (2020). https://doi.org/10.1186/s12877-020-01892-2
Speaker 5: Masato S. Abe (20 min)
Title: Behavioral complexity related to aging and cognitive decline
Abstract: Clarifying the relationship between brain systems and behavior is essential because it can lead to prediction and control of cognitive decline. In this presentation, I will introduce three topics on behavioral complexity related to aging and cognitive decline. First one is a relationship between word diversity and cognitive function in conversational data of the elderly people [1]. Second one is brain criticality hypothesis that a brain system sitting between disordered and ordered phases allows for flexible behavioral switching [2]. Third one is the loss of complexity of time series in daily activity related to aging.
References:
[1] Masato S. Abe, Mihoko Otake-Matsuura. Scaling laws in natural conversations among elderly people. PLOS ONE 16(2): e0246884 (2021). https://doi.org/10.1371/journal.pone.0246884
[2] Masato S. Abe. Functional advantages of Lévy walks emerging near a critical point. PNAS 117 (39) 24336-24344 (2020).https://doi.org/10.1073/pnas.2001548117
Speaker 6: Tomasz M. Rutkowski (20 min)
Title: Neurotechnology and Machine-learning Approaches to Early Dementia Onset Biomarker Development
Abstract: An increase in dementia cases is producing significant medical and economic pressure in many communities. This growing problem calls for applying AI-based technologies to support early diagnostics and subsequent non-pharmacological cognitive interventions and mental well-being monitoring. We present our recent efficient applications of machine learning (ML) methods concentrating on the `AI for social good’ application. We develop a digital dementia biomarker for early-onset dementia forecast. First, we will discuss two recent results demonstrating encouraging results of EEG-wearable-based signal analysis and subsequent classification. For the digital biomarker of dementia onset detection, we examine shallow- and deep-learning machine learning models. We report the best median accuracies in a range of 80~90% for random forest and fully connected neural network classifier models in both emotional face learning and reminiscent interior working memory paradigms. Next, we explain our second study results of conversational speech pattern-based prediction of mild dementia onset indicated by predictive Mini-Mental State Exam (MMSE) scores. Experiments with elderly subjects we conduct in natural conversation situations, with four members in each study group. We analyze the resulting four-party conversation transcripts within a natural language processing (NLP) deep learning framework to obtain conversation embedding. The best median MMSE prediction errors are at the level of 0.167, with a median coefficient of determination equal to 0.330 and a mean absolute error of 0.909. The reported outcomes showcase an essential social benefit of artificial intelligence (AI) employment for early dementia prediction. Furthermore, we improve ML practical utilization for the succeeding application in an uncomplicated and applied EEG-wearable examination or conversation analysis for dementia prediction.
References:
Rutkowski TM, Abe MS, Otake-Matsuura M. Neurotechnology and AI Approach for Early Dementia Onset Biomarker from EEG in Emotional Stimulus Evaluation Task. In: The 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE Engineering in Medicine and Biology Society. Virtual: IEEE Press; 2021. p. (accepted, in press).
Rutkowski TM, Abe MS, Komendzinski T, Otake-Matsuura M. Older Adult Mild Cognitive Impairment Prediction from Multiscale Entropy EEG Patterns in Reminiscent Interior Image Working Memory Paradigm. In: The 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE Engineering in Medicine and Biology Society. Virtual: IEEE Press; 2021. p. (accepted, in press).
Rutkowski TM, Abe MS, Tokunaga S, and Otake-Matsuura M. Dementia prediction in older people through topic-cued spontaneous conversation. medRxiv, 2021. https://doi.org/10.1101/2021.05.18.21257366
