Project Accomplishments: This project established the scientific, technical, and design foundation for the Digital Dyadic Coach (DDC)—a reinforcement learning (RL)-driven mobile intervention to optimize oral hygiene in dyads comprising older adults with Alzheimer’s Disease and Related Dementias (AD/ADRD) and their care partners. The project had two aims: developing a high-fidelity DDC through participatory design with older adults and care partners (Aim 1) andengineering a novel RL algorithm with dyadic functionalities that leverages real-time brushing data from Oral-B eBrush devices and mobile app interactions to adaptively deliver personalized digital prompts (Aim 2). The team enrolled 19 total participants (7 community-dwelling caregiver-care recipient dyads and 4 additional care recipients with a shared professional caregiver in memory care) after outreach to 520 potential participants, with 15 completing full study activities. 

Formative research including focus groups, think-aloud interviews, smart toothbrush use, and passive brushing monitoring revealed critical gaps in existing commercial oral health software—particularly the absence of dyadic synchronization between patient performance feedback and caregiver supportive prompting. These findings informed the development of a DDC conceptual model and low-fidelity wireframe prototype incorporating features such as music-based brushing cues, aquarium-based adherence visualization, smile/family selfie exchange, caregiver adherence summaries, and caregiver emotional check-ins. 

Initial Proposal Abstract: This project seeks to develop a novel AI-powered digital tool to empower older adults with Alzheimer’s disease and related dementias (AD/ADRD) to engage in oral self-care in at-home settings. More than a mere digital assistant, this Dyadic Digital Coach (DDC) will leverage the untapped potential of dyadic relationships between older adults and their primary caregivers.

Central to the DDC is the creation of a bespoke Reinforcement Learning (RL) algorithm specifically designed for dyadic interactions. The AI-driven approach will continually adapt to each participant’s unique needs in real-time, thereby optimizing behavioral intervention strategies and fostering positive care partner-patient interactions.

First, a usability study with an existing app will collect digital biomarkers of Oral Hygiene Practices (OHPs) by AD/ADRD participants. This data will drive a participatory design process aimed at developing a new app specifically for older adults with AD/ADRD and their care partners.

Second, we will develop a new dyadic RL algorithm that will learn and adapt the delivery of digital prompts to both the targeted older adult and their care partner. This will involve developing variants of the RL algorithm and comparing their performance via simulation studies.

Outcomes:

• 
  Oral Presentation: Digital Adaptive Interventions: Common Misconceptions and OpportunitiesMarch 25, 2026
  This presentation was given at the Prevention Research Center at Penn State University on March 25, 2026 (https://ssri.psu.edu/events/digital-adaptive-interventions-common-misconceptions-and-opportunities-prc-quantdev-joint) and at the Mobile Health Training Institute (mHTi) on April 13, 2026 (https://mhti.md2k.org/program/2026-program/).
• 
  Oral Presentation: Reinforcement Learning for Digital Health Interventions in the Dyadic SettingFebruary 4, 2026
  Abstract: We present our ongoing work on the development of an online reinforcement learning (RL) algorithm for dyadic digital intervention settings in which the task for the RL algorithm is to assist the target person with a difficult illness be adherent to behavioral activities. To achieve this goal the RL algorithm will not only… Read more: Oral Presentation: Reinforcement Learning for Digital Health Interventions in the Dyadic Setting
• 
  Oral Presentation: Just In Time Adaptive Interventions: Opportunities, Misconceptions, and the Promise of AI-based AlgorithmsJanuary 20, 2026
  Murphy, S.A., & Nahum-Shani, I. (2026, January) Just In Time Adaptive Interventions: Opportunities, Misconceptions, and the Promise of AI-based Algorithms. NIH Digital Health Scientific Interest Group (SIG)’s Speaker Series. NIH. Bethesda, MD.    You can watch this presentation here: https://videocast.nih.gov/watch/0b1c35f9-0852-11f1-9f14-124f0a52e769 
• 
  Publication: Replicable Bandits for Digital Health InterventionsJanuary 6, 2026
  Authors: Kelly W Zhang, Nowell Closser, Anna L Trella, Susan A Murphy Abstract Adaptive treatment assignment algorithms, such as bandit algorithms, are increasingly used in digital health intervention clinical trials. Frequently the data collected from these trials is used to conduct causal inference and related data analyses to decide how… Read more: Publication: Replicable Bandits for Digital Health Interventions
• 
  Oral Presentation: Online Reinforcement Learning in Digital Health InterventionsOctober 25, 2025
  Abstract: In this talk, Susan will discuss first solutions to some of the challenges we face in developing online RL algorithms for use in digital health interventions targeting patients struggling with health problems such as substance misuse, hypertension, and bone marrow transplantation. Digital health raises a number of challenges to… Read more: Oral Presentation: Online Reinforcement Learning in Digital Health Interventions
• 
  Publication: Just-in-Time Adaptive Interventions: Where Are We Now and What Is Next?September 5, 2025
  Authors: Inbal Nahum-Shani, Susan A Murphy Abstract The past decade has seen a surge in developing just-in-time adaptive interventions (JITAIs)-an intervention approach that leverages advancements in digital technologies to address the rapidly changing needs of individuals in daily life. This article provides an overview of the state of science on… Read more: Publication: Just-in-Time Adaptive Interventions: Where Are We Now and What Is Next?
• 
  Publication: Reinforcement Learning on Dyads to Enhance Medication AdherenceJune 23, 2025
  Authors: Ziping Xu, Hinal Jajal, Sung Won Choi, Inbal Nahum-Shani, Guy Shani, Alexandra M. Psihogios, Pei-Yao Hung, Susan A. Murphy Abstract Medication adherence is critical for the recovery of adolescents and young adults (AYAs) who have undergone hematopoietic cell transplantation. However, maintaining adherence is challenging for AYAs after hospital discharge,… Read more: Publication: Reinforcement Learning on Dyads to Enhance Medication Adherence
• 
  Publication: Causal Directed Acyclic Graph-informed Reward DesignJune 3, 2025
  Authors: Luton Zou, Ziping Xu, Daiqi Gao, Susan Murphy Abstract It is well known that in reinforcement learning (RL) different reward functions may lead to the same optimal policy, while some reward functions can be substantially easier to learn. In this paper, we propose a framework for reward design by… Read more: Publication: Causal Directed Acyclic Graph-informed Reward Design
• 
  Publication: Digital Twins for Just-in-Time Adaptive Interventions (JITAI-Twins): A Framework for Optimizing and Continually Improving JITAIsMay 9, 2025
  Authors: Asim H. Gazi, Daiqi Gao, Susobhan Ghosh, Ziping Xu, Anna Trella, Predrag Klasnja, Susan A. Murphy Abstract Just-in-time adaptive interventions (JITAIs) are nascent precision medicine systems that extend personalized healthcare support to everyday life. A challenge in designing JITAIs is that personalized support often involves sophisticated decision-making algorithms. These… Read more: Publication: Digital Twins for Just-in-Time Adaptive Interventions (JITAI-Twins): A Framework for Optimizing and Continually Improving JITAIs
• 
  Poster Presentation: a2 National SymposiumApril 4, 2025
  Title: A Digital Dyadic Coach to Promote Oral Health Self-Care in Older Adults Authors: Guy Shani, Vivek Shetty, Jenin Alcaraz, Susan A Murphy, Inbal Billie Nahum-Shani