In partnership with
Effortless Tutorial Video Creation with Guidde
Transform your team’s static training materials into dynamic, engaging video guides with Guidde.
Here’s what you’ll love about Guidde:
1️⃣ Easy to Create: Turn PDFs or manuals into stunning video tutorials with a single click.
2️⃣ Easy to Update: Update video content in seconds to keep your training materials relevant.
3️⃣ Easy to Localize: Generate multilingual guides to ensure accessibility for global teams.
Empower your teammates with interactive learning.
And the best part? The browser extension is 100% free.
Hi!
Welcome to AIMedily.
This week, I came across a paper in the The New England Journal of Medicine that discussed the impact of AI in rural communities.
It highlighted something many of us have seen: nurses in remote areas caring for dozens of patients spread across hundreds of miles, often with limited support.
The authors outlined how AI could help them triage, respond to emergencies faster, and reduce documentation load so they can focus on patient care.
It brought me back to the year I spent as a general practitioner in a small town during medical school. I had two nurses in the morning, and after that, I was on my own, even at night. Some days, there were long lines of patients waiting. Other days, I had to refer someone to a hospital hours away.
Reading this paper, I could clearly imagine how AI —if built safely and responsibly—could have helped me deliver better care back then.
We’re not fully there yet, but I hope AI will become a way to strengthen healthcare for rural communities.
Let’s dive into today’s issue.
🤖 AIBytes
Researchers developed a shared-control bionic hand that blends human intent and machine learning to improve precision, reduce effort, and allow safe interaction with different objects.
🔬 Methods
5 adults with upper-limb loss used a robotic bionic hand with 27 degrees of freedom.
Muscle signals from the forearm (surface electromyography) tracked the user’s intention.
The shared control system predicted the shape of the hand needed for the task using learned patterns from human grasping.
Users performed several tests, including:
Grasping objects
Matching target hand postures
Moving objects in functional tasks
Researchers compared:
User-only control
Shared human–machine control
📊 Results
The shared system reduced finger-angle errors by more than half compared with user-only control.
Movements were faster, with smoother transitions.
Precision tasks had higher accuracy.
The system helped users avoid unstable or incorrect hand shapes.
Functional tests showed more reliable grasping, including small and irregular objects.
Users reported the movements felt more intuitive and required less effort.
🔑 Key Takeaways
Shared human–machine control made bionic hand movement better and more precise.
The system predicted the target hand shape, reducing strain and error.
Allowing humans to modulate force is essential for safety and real-world function.
This approach improved handling of small and complex objects.
💡Shared control is a promising direction for prosthetics, enabling precision, adaptability, and reduced effort.
🔗 Zhang A, et al. Shared human–machine control enables intuitive and precise multi-digit manipulation with a bionic hand. Nature Communications. 2025. DOI: 10.1038/s41467-025-65965-9.
This randomized clinical trial tested whether two ambient AI scribes could safely reduce documentation time and improve physician well-being in real outpatient practices.
🔬 Methods
Participants: 238 outpatient physicians from 14 specialties.
Usual-care control (no AI scribe)
Primary outcome: Change from baseline in log writing time-in-note (Epic software).
📊 Results
Use of AI scribes:
DAX: Used in 33.5% of 24,696 visits.
Nabla: Used in 29.5% of 23,653 visits.
Documentation time:
Nabla: 9.5% decrease significantly in time-in-note versus control.
DAX Copilot did not produce a statistically significant reduction.
Burnout and workload:
Changes suggest improvement in burnout, task load, and work exhaustion when using AI scribes, compared to control.
Safety, inaccuracies, and bias:
Clinically significant inaccuracies were reported “occasionally”.
One grade 1 adverse event was reported.
Biases were uncommon but still present.
🔑 Key Takeaways
Nabla reduced documentation time. DAX did not show a significant change compared with usual care.
Both AI scribes may improve burnout measures, task load, and work exhaustion..
Clinically meaningful inaccuracies can still occur. Most often as omissions or subtle errors, so physicians must stay vigilant.
AI scribes are a promising tools, but there is a need for larger, multicenter studies and continuous monitoring of safety and bias.
🔗 Lukac PJ, Turner W, Vangala S, et al. Ambient AI Scribes in Clinical Practice: A Randomized Trial. NEJM AI. 2025;2(12). doi: 10.1056/AIoa2501000.
🦾TechTool
Generates realistic video clips from a text prompt.
Handles complex motion well, making it useful for presentations.
Produces clean, professional footage that’s easy to edit.
Uses an Emotionally Intelligent Voice Interface that reads tone.
Adapts responses in real time based on the user’s emotional state for more natural interactions.
You can create audio books, podcasts, and conversational agents.
Biomechanical motion-capture and analysis for gait, posture, and movement assessments.
Translates raw data into clear reports and visual summaries.
Supports outcome measurement over time, useful for tracking progress or comparing interventions.
🧬AIMedily Snaps
An AI tool that identify undiagnosed Alzheimer's cases while reducing disparities (Link).
What will be the first AI-designed drug? These disease-fighting antibodies are top contenders (Link).
AI technologies that can shape the future of health care (Link).
Stanford researchers work to squash ‘fantastic bugs’ hidden in AI benchmarks (Link).
U.S. Department of Health and Human Services adopts AI (Link).
Microsoft joins the ambient AI scribes game for ambulatory providers (Link).
🧩TriviaRX
Which physiological signal was the first to be analyzed with a machine-learning algorithm in scientific research (1960)?
A) Electroencephalography (EEG)
B) Electromyography (EMG)
C) Electrocardiography (ECG)
D) Respiratory effort signals
Now, last week’s TriviaRX Answer
Which specialty has the most FDA-cleared AI/Machine Learning medical devices?
✅ B) Radiology
Radiology holds the largest share of FDA-cleared AI/ML medical devices.
That’s it.
Thank you for taking the time to read.
You’re already ahead of the curve in medical AI — don’t keep it to yourself. Forward AIMedily to your colleagues who’d appreciate the insights.
Until next Wednesday.
Itzel Fer, MD PM&R
Forwarded this email? Sign up here
P.S. Enjoying AIMedily? 👉 Write a review here (it takes less than a minute).
P.S. I’ve been trying to find a photo from my time in that small town I mentioned. If I find one, I’ll share it with you.
