Attention is all you need, but misdirected. Sharing my new curmudgeon corner piece at AI & Society reflecting my thoughts on emerging societal issue on our attentions rdcu.be/eMLDy

EXCITED TO SHARE! Here we show synthetic clinical data may be used in place of real-world data to answer research questions, train ML models, and to amplify sample sizes Check it out: thejns.org/focus/view/j... @emreszg.bsky.social

🚨 New paper alert We analyzed 113K+ interactions from 24K+ users of a mHealth conversational agent (CA) used for self-care & chronic disease management. 📱🤖 For more details: link.springer.com/article/10.1... #DigitalHealth #ConversationalAI #mHealth #HealthInformatics #ChronicDiseaseManagement

Tomorrow at CCHEOR C.A.R.E.S, I’ll present new results from our study on an AI chatbot helping low-income families to navigate social needs land connect with community resources. Register: nationwidechildrens.cloud-cme.com/course/cours... @nationwidekids.bsky.social #DigitalHealth #conversationalAI

Wrapping up an inspiring week at #SBM2025, connecting w/ researchers driving innovation in behavioral science & digital health. Also, I was honored to receive the SBM Digital Health SIG Award for Excellence in Digital Health Publication—a humbling recognition + motivation to keep up the momentum!

It is poster time! Find me at D77 to learn more about our chatbot for helping caregivers of young childhood survivors at #SBM2025 @sbm.org

On my way to #SBM2025 I’ll be presenting our work on using conversational AI to support caregivers of childhood cancer survivors.

New controlled study of AI voice mode on mental health finds complex results: GPT-4 with an empathetic voice reduced loneliness, but heavier users, and especially when use was combined with a more neutral voice or those with particular personalities, had multiple negative impacts on users.

I have a knack for understanding if a human or a machine is "particularly impressed by" my recent article... #academicLife

Super excited to share our latest preprint: We generate synthetic neurosurgical data with an LLM (using a single prompt) and train an ML model on it to predict real-world neurosurgical outcomes Read more: arxiv.org/pdf/2502.09566 @emreszg.bsky.social

Excited to share the first in our series of papers showcasing the ability to generate synthetic clinical datasets with LLMs, using zero-shot prompting Read more: www.frontiersin.org/journals/art... @emreszg.bsky.social

Pleased to share our recent publication @jmirpub.bsky.social. In this work, my colleague @abakik.bsky.social and I explore how generalist conversational AI (CAI), powered by large language models, can transform public health communication. www.jmir.org/2025/1/e69007

🔔 Our new manuscript is out- seeking for an answer to the question: 🤖 Can large language models (LLM) and artificial intelligence help addressing the information needs of pediatric cancer caregivers ❓ A thread 🧵 onlinelibrary.wiley.com/doi/full/10....

With chatGPT on WhatsApp and phone calls, as well as meta AI, My colleague @abakik.bsky.social and I have been discussing the huge potential of such integration of generalist conversational AI to current messaging services in public health. More details are on our preprint @jmirpub.bsky.social

Addressing #AI bias and lack of transparency for healthcare datasets Consensus recommendations from a global consortium thelancet.com/journals/lan...

As AI generated content has been used all around our digital ecosystem, I started to appreciate typos a bit more, an indication of human touch, until someone prompt chatgpt to add typos...

🚨 A new study raises a key concern: AI designed to ease radiologists' workload might increase burnout instead. jamanetwork.com/journals/jam... #HealthTech #AIEthics #ProfessionalWellbeing #Burnout #AI

The New York Times has once again chosen a provocative headline, a common but risky strategy they use to capture attention. While this approach may drive clicks, it risks fostering an imbalanced trust in AI at the expense of healthcare providers.

Our collaborative work with DiMe members explores current research landscape reporting sensor-based digital health technologies (sDHTs) human factors, human-centered design, and usability. Available at JMIR 👉 jmir.org/2024/1/e57628/

I am happy to share our first publication from DAPHNE project! Our study highlights the design and evaluation of our DAPHNE chatbot, aimed at screening and understanding social needs and linking patients and families to essential resources. humanfactors.jmir.org/2024/1/e57114/ #digitalhealth