Samir Akre, PhD

Unveiling the Complex Relationship Between Sleep, Depression, and Neurocognitive Performance Self-reported and physiological sleep data reveal different insights into depression. Our study uncovers surprising gaps between subjective sleep experiences and objective measures, highlighting the n...

While wearables offer promising tools, they’re not yet a replacement for self-reports. Instead, they highlight the complexity of mental health—where both subjective experiences and objective data have unique roles to play.

Check out the "Behind the paper" blog post here:
go.nature.com/4jLA6x5

This raises critical questions:
• What do subjective vs. objective sleep measures really capture?
• Can we ever rely solely on passive data to understand mental health?
• How should we combine these data sources for better insights?

In our latest study, we find that how participants self-report their sleep quality differs dramatically from their wearable device measured sleep – and that both were important to understanding their depression symptoms.

Huge thank you to all the co-authors!@depressionlab.bsky.social

Comparing self reported and physiological sleep quality from consumer devices to depression and neurocognitive performance npj Digital Medicine - Comparing self reported and physiological sleep quality from consumer devices to depression and neurocognitive performance

𝗪𝗵𝗮𝘁 𝗖𝗮𝗻 𝗪𝗲𝗮𝗿𝗮𝗯𝗹𝗲𝘀 𝗧𝗲𝗹𝗹 𝗨𝘀 𝗔𝗯𝗼𝘂𝘁 𝗦𝗹𝗲𝗲𝗽 𝗶𝗻 𝗠𝗲𝗻𝘁𝗮𝗹 𝗛𝗲𝗮𝗹𝘁𝗵? 🧠⌚ 💤
New paper just dropped in npj Digital Medicine!
rdcu.be/d9b1b

I entered this research thinking wearable devices could eventually replace self-reported measures for sleep quality in mental health studies – but it seems I was wrong.

Advancing digital sensing in mental health research - npj Digital Medicine npj Digital Medicine - Advancing digital sensing in mental health research

Awesome new paper by @samirakre.com and colleagues on "Advancing digital sensing in mental health" discussing challenges around data standards & translational impact + proposing strategies to address these challenges to ensure rigorous, equitable, & impactful research. www.nature.com/articles/s41...

This was one of my favorite papers to get to author during my PhD. Such an honor to put together the thoughts from many of the people I respect most in the field.

3) Anomaly detection doesn't seem to be as important for depression symptoms as it is for other conditions like relapse in those with schizophrenia (at least in a 3 month timescale)

4) We can kind of detect momentary subjective state (reward & affect) with just 2 hours of passive data!

Key points:
1) We can indeed detect some *specific* symptoms of depression with passive wearable data and no prior information from a given individual (nomothetic modeling)

2) self-reported and passively measured sleep quality are different in how they relate to depression, both are important

Machine Learning-Based Detection of Depression Symptoms with Smartphones and Consumer Wearable Devices Author(s): Akre, Samir | Advisor(s): Bui, Alex AT | Abstract: Consumer wearable devices and smartphones are ubiquitous and generate valuable health-related data that remain under explored. These data ...

Woo my PhD dissertation "Machine Learning-Based Detection of Depression Symptoms with Smartphones and Consumer Wearable Devices" is finally public.

Let me know if you want to chat about it!

See 🧵 for some takeaways
escholarship.org/uc/item/04p4...

Thanks! I may just need to replace all my poetry based packages with uv...

Whats the transition like from poetry?

I didn't know I was looking for this content