Hallo…
Today’s AI research findings are mostly optimistic (except #10). I’m personally excited to see how AI is starting to improve healthcare and even save lives. Of course, there are still challenges and important questions ahead, but this issue focuses mainly on positive and practical ways AI is being used in health and healthcare.
Here are today´s 10 insights, from fist bump scans to future medicine and everything in between.
#1
〰️ Diagnosing disorders from a fist bump
I love the idea that a picture of your hand could be more revealing than a long talk with a specialist. Kobe University researchers taught an AI to spot a rare hormonal disorder called acromegaly just by looking at photos of the back of a hand or a clenched fist. 👊🏼 The AI is actually better at spotting the disease than expert human doctors, and it protects patient privacy because it doesn’t even need to see your face. → More
#2
〰️ A liquid biopsy to catch liver disease early ✅
Imagine a blood test that can scan 40 million genomic patterns at once to tell you if your liver is in trouble. Johns Hopkins researchers built an AI that looks at DNA fragments like a detective, catching liver fibrosis early enough to actually reverse it before it turns into cancer. It's like having a high-tech early warning system inside your veins. → More
#3
〰️ AI finds the needles in the medical haystack
Cleveland Clinic used an AI called Synapsis to do the impossible: it screened nearly 1,500 patient records in just one week to find candidates for a rare heart disease trial. It found 30 perfect matches, and human recruiters had missed 29 of them. It’s basically a super-powered research assistant that ensures no patient gets left behind. 💚 → More
#4
〰️ AI Makes Lung Radiation Easier for Patients
Mapping out the lobes of a human lung for radiation therapy usually requires patients to hold their breath, which is obviously hard to do when you’re sick. Researchers created an AI tool that can accurately map lungs even while the patient is breathing normally. It’s a huge win for clinical consistency and making a scary treatment just a little bit easier. ✅ → More
#5
〰️ AI-discovered drugs are officially hitting human trials
We’ve been hearing about AI-designed drugs for years, but now it’s actually happening. A new drug for kidney disease anemia, discovered by an AI platform called Chemistry42, has just started Phase I human trials. It took a fraction of the usual time to go from a computer idea to a real pill, proving that AI can seriously speed up life-saving medicine. I look forward to hearing how this turns out 🤔 → More
#6
〰️ The Medical GPS for Doctors
Medical data now doubles about every 73 days, leaving doctors buried in huge guideline manuals. A €9.5M project called GUIDE‑AI is building digital assistants inside hospital software that scan patient data in real time. The system works like a medical GPS, warning doctors if a treatment plan drifts from the newest guidance for diseases like asthma and heart failure. It helps clinicians stay current while they still make the final call. 👩🏼⚕️👨🏾⚕️→ More
#7
〰️ Traditional Chinese Medicine gets a digital 2.0 upgrade
How do you turn ancient wisdom into data?. Researchers are using something called TCM Phenomics 2.0 to integrate Traditional Chinese Medicine with AI and wearable devices. By using algorithms to analyze things like tongue appearance and pulse, they are moving TCM from gut feelings to precise, evidence-based science. → More
#8
〰️ AI predicts your baby’s birthday from a single photo
Standard ultrasound dates can be a bit of a guessing game, but the FDA just cleared a new tool that uses AI to predict a baby’s due date from standard images. Trained on millions of scans, it cuts through the uncertainty of pregnancy dating, which could be a literal life-saver for preventing preterm births. 🚼 👶🏼 → More
#9
〰️ Canada’s new Super-Brain for health data
Canada just launched ARCHIMEDES, a national platform that lets researchers safely peek at health data from all over the country. Instead of data being stuck in silos at different hospitals, this platform brings it all together so AI can find trends in everything from heart disease to mental health. I wonder what the implications would be ⍰ → More
#10
〰️ The scary Model Collapse of medical records
If we start letting AI write all our medical notes, we might be in big trouble. 🚨Researchers found that if a new AI is trained on notes written by a previous AI, critical details like lung mass or bone fracture start to disappear after just two generations. It’s a warning that we need to keep human data in the loop to keep medical records from becoming clinically useless. → More
That´s it for today…Thank you for reading!
What is one thing that struck you from today´s insights?
If you know anyone who would like to read these insights, please share this newsletter with them.
Stay curious and in the loop.