Year 3 Pilots

Publication: Replicable Bandits for Digital Health Interventions

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 to refine the intervention, and whether to roll-out the intervention more broadly. This work studies inference for estimands that depend on the adaptive algorithm itself; a simple example is the mean reward under the adaptive algorithm. Specifically, we investigate the replicability of statistical analyses concerning such estimands when using data from trials deploying adaptive treatment assignment algorithms. We demonstrate that many standard statistical estimators can be inconsistent and fail to be replicable across repetitions of the clinical trial, even as the sample size grows large.…

Continue ReadingPublication: Replicable Bandits for Digital Health Interventions

Oral Presentation: Can You Walk Me Through It? Explainable SMS Phishing Detection using LLM-based Agents

This is part of the monthly MassAITC webinar series. Abstract: Phishing attacks pose a significant threat to users, especially older adults. Existing defenses mainly focus on phishing detection but often cannot explain to lay users why a message is malicious. In this talk, I will discuss how we use Large Language Models (LLMs) to detect SMS phishing while generating evidence-based explanations. The key challenge is that SMS is short, lacking the necessary context for security reasoning. We develop a prototype called SmishX that gathers external contexts to augment the chain-of-thought (CoT) reasoning of LLMs and facilitate the explanation process. I will further discuss our user studies to evaluate the effectiveness and usability of SmishX. Finally, I will discuss the open challenges and opportunities of using AI to help older adults better protect themselves from cybersecurity…

Continue ReadingOral Presentation: Can You Walk Me Through It? Explainable SMS Phishing Detection using LLM-based Agents

Oral Presentation: American Speech Language Hearing Association Conference – 2025

SLPs on the Line: Engaging Patients in Cognitive Rehabilitation via Telephone: Moneta Health has developed a telephone-based telepractice model to deliver cognitive rehabilitation therapy (CR) to older adults with cognitive impairment. This model was designed to address common barriers to accessing quality care through traditional in-clinic and telehealth services. With Moneta, patients receive telephone sessions delivered by speech-language pathologists (SLPs), and sessions delivered by an AI-powered automated agent. These automated sessions contain personalized, interactive cognitive activities designed and selected by an SLP. An analysis of over 100 patients who completed the program shows high engagement, compliance and satisfaction, supporting the use of digital and audio-only delivery to promote a positive patient experience. In this paper, we provide an overview of the patient experience with Moneta, and review engagement and satisfaction metrics. Source: https://plan.core-apps.com/asha2025/event/28734884ce4484665e296583a03904a0

Continue ReadingOral Presentation: American Speech Language Hearing Association Conference – 2025

Award: Best Demo Audience Choice Award at IEEE BSN 2025

The team of Colin Barry, Tatsuo Kumamoto, Edward Wang, and Lina Battikha won the Audience Choice Award for Best Demo at the IEEE-EMBS International Conference on Body Sensor Networks – Computational Medicine: Expanding Health through Sensing and AI held from November 3-5, 2025 for their demo entitled, "Oscillometric Smartphone Blood Pressure Demo." Source: LinkedIn Post

Continue ReadingAward: Best Demo Audience Choice Award at IEEE BSN 2025

Oral Presentation: Online Reinforcement Learning in Digital Health Interventions

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 the RL community including different sets of actions, each set intended to impact patients over a different time scale; the need to learn both within an implementation and between implementations of the RL algorithm; noisy environments; and a lack of mechanistic models. In all of these settings, the online line algorithm must be stable and autonomous. Despite these challenges, RL, with careful initialization, with careful management of bias/variance tradeoff, and by close collaboration with health scientists, can be successful. We can make an impact! Dr. Susan Murphy has…

Continue ReadingOral Presentation: Online Reinforcement Learning in Digital Health Interventions

Funding: $4.5M seed round raised

Seed round raised from existing investors (True Ventures, BKR Capital, Centre for Aging and Brain Health Innovation, Health2047 and others) - $4.5M to date. Securing of MassAITC pilot funding helped in due diligence with investors for their seed round, and thus was critical to closing their funding including with Health2047, venture arm of the American Medical Association and with other early stage investors. Source: businesswire

Continue ReadingFunding: $4.5M seed round raised