In partnership with
Find out why 1M+ professionals read Superhuman AI daily.
AI won't take over the world. People who know how to use AI will.
Here's how to stay ahead with AI:
Sign up for Superhuman AI. The AI newsletter read by 1M+ pros.
Master AI tools, tutorials, and news in just 3 minutes a day.
Become 10X more productive using AI.
Hi!
Welcome back to AIMedily.
This week, I want to start thanking my editor: my 15-year-old daughter.
Every week, before I click send, she reads and polishes each issue.
As you may know, English isn’t my first language — Spanish is — and while I feel comfortable writing in English, her thoughtful edits make this newsletter clearer and better.
I’m looking to feature some testimonials from readers like you on the AIMedily website. 2 or 3 sentences are enough to highlight what you enjoy most about AIMedily.
Thank you for your consideration. Here is the link. Fill review here
Now, time to dive into today’s newsletter.
🤖 AIBytes
This study developed and validated Machine Learning models to predict independent walking in children with Cerebral Palsy (CP).
🔬 Methods
Study design: Retrospective cohort.
Participants:
807 children with CP (aged 0–6 years) from the Henan CP Registry Platform. The patients were followed up by telephone.
Predictors used to train the Machine Learning models:
Neurodevelopmental milestones: Sitting, standing, and walking.
Gross Motor Function Classification System (GMFCS) before age 2.
Gross Motor Function Measure 88 (GMFM-88).
Gestational age at birth
Neonatal asphyxia
Hyperbilirubinemic encephalopathy
Cerebral Palsy subtype
MRI classification: white or grey matter injury, brain malformations, focal vascular injury, miscellaneous patterns, and normal imaging.
Comorbidities and interventions: epilepsy, visual or intellectual disability, and surgeries.
📊Results
69.5% (561/807) of children achieved independent walking by age 6.
Top predictors:
GMFCS measures-88 level before age 2.
Independent sitting by age 2.
The next most important predictors were:
Cerebral Palsy subtype: hemiplegia had the highest walking rate with 93.96%
76.57% for diplegia, 25.49% for dyskinetic, and 22.32% for quadriplegia
MRI classification
Intelectual Disability
Early preterm birth
🔑 Key Takeaways
Machine learning models can predict walking ability in Cerebral Palsy.
Sitting by age 2 and GMFCS level were the strongest predictors for independent walking.
💡This model has the potential to be a helpful tool for early prognosis. It can also help plan rehabilitation and give the family orientation.
🔗Wang Y, Yang Y. Development and Validation of a Prognostic Model for Independent Walking in Children With Cerebral Palsy Based on Machine Learning. Arch Phys Med Rehabil. 2025;000:1–9. https://doi.org/10.1016/j.apmr.2025.05.006
This study evaluated a wearable robotic hand exoskeleton (HandMATE) for stroke patients.
They did a prospective pilot study to test the robot in unsupervised home-based rehabilitation.
🔬 Methods
The study had two phases: clinician-supported and unsupervised treatment.
Participants: 14 chronic stroke survivors (10 female, 4 male) with moderate hand impairment (average Fugl-Meyer 24). 10 completed the final evaluations.
HandMATE: a wearable exoskeleton that was personalized to assist based on the user's intent to move the fingers (see image of the robot here).
Intervention: HandMATE exoskeleton used at home:
Phase 1 (1 month): Home use with weekly clinical visits.
Phase 2 (3 months): Continued home use without supervision.
Assessments:
📊 Results
Phase 1 (3.8 hours of use per week):
FMA improved by +3.69 points (p = 0.0027)
ARAT improved by +1.85 points (p = 0.0346)
Phase 2 (1.9 hours of use per week): Improvement continued but was not statistically significant:
FMA +2.67 (p = 0.0903)
ARAT +2.67 (p = 0.0903)
Range of motion improved significantly after both phases.
User feedback: High satisfaction; usage declined over time without clinician supervision.
🔑 Key Takeaways
The HandMATE exoskeleton significantly improved hand function and range of motion for patients with chronic stroke.
The use and consistency dropped when patients were not being supervised.
Adding gamification and engagement tools is necessary to keep patients interested.
💡Exoskeletons like HandMATE, have the potential to support home-based treatment and lower cost barriers in post-stroke.
🔗Pergami-Peries M, Grainger M, Lum PS. HandMATE: Advancing Accessible Hand Rehabilitation for Persons With Stroke. IEEE Trans Neural Syst Rehabil Eng. 2025;33(8):3449–3457. doi:10.1109/TNSRE.2025.3528987
🦾TechTool
ReHUB: The patient can perform the prescribed exercises from their mobile phone, tablet, or computer and receive real-time visual and auditory feedback for a correct execution of the exercise. It also reports information to the medical team.
Explain Paper: With this platform, research papers become easy to understand. You can upload a paper, highlight the area that has a complex concept you need help understanding, and it will give you a clearer explanation.
Grammarly AI: You probably know (or use) this platform. They recently launched AI agents that can predict readers’ questions, strengthen your argument, grade your work, and generate proper citations, besides checking your grammar.
Sensoria-mat: A Wheelchair Cushion with textile pressure sensors to track and monitor the number of front leans, side-to-side leans, and push-ups completed to prevent pressure ulcers. It sends notifications to the user to remind them to relieve pressure. If you want to read more about it, here’s a paper Link).
🧬AIMedily Snaps
Cleveland Clinic and Piramidal.ai will test an AI model in the Intensive Care Unit that can analyze electroencephalograms in real-time and provide interpretation and insights (Link).
How Google and NASA are testing AI for medical care in space (Link).
Here is the guide from the American Medical Association for health care systems to implement, manage, and scale AI (Link).
🧩TriviaRX
In what year was the term “electromyography” (EMG) first officially used in published medical literature?
A) 1890
B) 1948
C) 1902
D) 1975
Now, the correct answer from last week's TriviaRX 🥁
✅B) Whale baleen 🐳
In the 18th and 19th centuries, whale baleen (the flexible keratin plates from whales’ mouths) was used to stiffen orthopedic body braces and posture-correcting corsets.
That’s it for today. Thanks for reading.
If you found this email useful, feel free to forward this email to a friend or colleague (I will be forever grateful 🥰).
Thank you!
Until Friday with updates on Large Language Models.
Itzel Fer, MD PM&R
Forwarded this email? Sign up here
P.S. And if you gave a review, THANK YOU!
How did you like today's newsletter?
Fact-based news without bias awaits. Make 1440 your choice today.
Overwhelmed by biased news? Cut through the clutter and get straight facts with your daily 1440 digest. From politics to sports, join millions who start their day informed.
1
