Sponsored by
The most trustworthy AI admin agent for busy executives.
Busy executive. Packed calendar. An inbox that never empties.
Everyone is talking about AI agents. Every week brings another promise of an assistant that can do it all.
Catch is the real deal.
A smart, proactive AI admin agent focused solely on taking administrative work off your plate.
It schedules meetings, triages your inbox, drafts emails in your voice, resolves conflicts, sends follow-ups, and handles the countless small tasks that consume your day.
Available wherever you work — Gmail, Outlook, Slack, WhatsApp, and even over the phone.
No setup. No training. No learning curve.
Catch learns how you work, takes action when it's confident, and keeps things moving without constant supervision.
From swamped to sorted in seconds.
Get started with Catch and have your assistant ready before your next meeting.
Hi!
Welcome back to AIMedily.
I spent last week at Stanford for the AIMI and RAISE Health Symposiums.
Being there reminded me how much is happening in medical AI right now, and how important it is for clinicians to stay close to these conversations.
I had the chance to meet researchers from leading medical institutions (I was super excited about this), and also people working on this from companies like OpenEvidence, Google DeepMind, and OpenAI.
I left grateful for the conversations, the ideas, and the chance to keep learning from people who are thinking deeply about this work.
I’m excited to keep sharing what I learn with you through AIMedily.
Let’s dive into today’s issue.
🤖 AIBytes
Researchers built Pathog-PDx, a machine learning system that uses routine clinical data to predict respiratory pathogens in hospitalized children.
The goal was to help clinicians identify likely pathogens before culture or nucleic acid test results are ready.
🔬 Methods
Study type: Large multicenter model development and validation study
Patients: 134,500 hospitalized children
Final model cohort: 85,349 children
Prospective test cohort: 1,338 children
Data used: 42 routine clinical and lab features from electronic health records
Model task: Predict infection site, 22 pathogen subtypes, mixed infection, and pediatric ICU risk
Validation: Internal, external, and prospective validation cohorts
📊 Results
The model helped identify the likely infection site and pathogen type from routine clinical data.
It could distinguish bacterial, viral, and fungal infections, and also helped detect mixed infections.
For influenza, it correctly identified 88% of true cases and correctly ruled it out in 86% of children without influenza.
In the prospective cohort of 1,338 children, Pathog-PDx gave predictions within 3–6 hours after admission.
By comparison, nucleic acid tests took about 20 hours, and cultures took about 3.5 days.
Useful as an early decision-support tool, not as a replacement for confirmatory testing.
🔑 Key Takeaways
Pathog-PDx may help clinicians narrow the likely cause of pediatric respiratory infection earlier.
The model uses routine clinical and lab data.
It may be useful when confirmatory results are delayed or to narrow diagnosis.
External testing showed lower performance, so local validation is still needed before clinical use.
🔗 Su D, Chen Q, Xu R, et al. Development and validation of a machine learning-based diagnostic system for 22 pediatric respiratory pathogens: a large-scale multicenter study. npj Digital Medicine. 2026. https://doi.org/10.1038/s41746-026-02818-9
Researchers studied whether a large cancer screening trial using a cell-free DNA multicancer early detection test was linked with longer diagnostic delays in England.
The question was not only whether the test could find cancer. It was whether the health system could handle the extra demand.
🔬 Methods
Design: Cross-sectional study using difference-in-differences analysis
Setting: 21 cancer regions in England
Comparison: 8 regions in the NHS-Galleri trial vs 13 regions not in the trial
Test: A cell-free DNA (cfDNA) multicancer early detection test that uses AI to screen for more than 50 cancer types.
Data: 9.6 million suspected cancer referrals
Main outcome: Diagnostic delay, defined as taking more than 28 days to reach diagnostic resolution
📊 Results
The study included 1,875,236 referrals for suspected head and neck, lung, or upper gastrointestinal cancers.
In the first 6 months, diagnostic delays increased in trial regions from 28.6% to 29.6%.
In non-trial regions, delays decreased from 28.9% to 26.3%.
Trial participation was linked with a 3.4 percentage-point increase in diagnostic delays (P < .001).
The increase continued during the second 6-month period: 4.8 percentage points (P = .003).
After the first year, the difference was no longer statistically significant.
The authors estimated 9,591 additional referrals experienced diagnostic delay during the first year.
Average wait time to diagnostic resolution increased by about 2 days in participating regions.
🔑 Key Takeaways
Finding more possible cancers is only helpful if the health system can evaluate them quickly.
In this study, regions in the MCED trial had a temporary rise in cancer diagnostic delays.
The average delay was about 2 extra days, and was unlikely to change the trial’s main cancer-stage results.
🔗 Mann S, Nascimento de Lima P, Eagan J, Ulyte A, Griffin BA. Cancer Diagnostic Delay Rates Associated With a Population-Based Screening Trial Evaluating a Cell-Free DNA Multicancer Early Detection Test. JAMA. Published online May 30, 2026. https://doi.org/10.1001/jama.2026.6803
🦾TechTools
Analyze ECGs and detect patterns that may not be obvious on routine review.
Its tools support ECG interpretation, disease risk scoring, and remote ECG consultation.
Useful for earlier risk detection.
AI platform used to support radiology and acute care workflows.
It can flag urgent findings, help prioritize imaging, and connect care teams faster.
Helps identify who needs attention and who needs follow-up.
📈 Productivity Tool of the Week:
Connects apps so repetitive tasks can happen automatically.
You can use it to move information between forms, spreadsheets, email, calendars, and other tools.
Useful for reducing admin work.
🧬AIMedily Snaps
Nature: How good are ‘AI doctors’ — and will they take over medicine? (Link).
CHAI: Coalition for Health AI Releases Comprehensive Governance Playbooks to Streamline AI Implementation for Health Systems (Link)
Wolters Kluwer: Rapid AI adoption in healthcare highlights worries, opportunities, for both patients and clinicians (Link).
AMA: Millions use chatbots for mental health. Accountability can’t wait (Link).
Anthropic: released their new models Fable 5 and Claude Mythos 5 (Link).
White House: Trump Executive Order Addresses Cybersecurity with Voluntary Framework for AI (Link).
🧪Research Signals
NEJM: MEDS — An Emerging Data Standard and Ecosystem for Health AI Research (Link).
Nature: ChatGPT in the diagnosis and management of complex polyneuropathies: comparative analysis with neurologists using real-world cases (Link).
Nature: Human–AI collaboration for dysphagia rehabilitation from effectiveness to implementation complexity: a systematic review (Link).
JMIR: Ambient AI Scribes to Create Educational Feedback Notes for Medical Students: Randomized Trial (Link).
Nature: A clinical neuroimaging platform for rapid, automated lesion detection and personalized post-stroke outcome prediction (Link).
Nature: A large-scale vision foundation model for musculoskeletal radiographs (Link).
🧩TriviaRX
Before modern home pregnancy tests, which animal was once used as a living pregnancy test?
A) Rabbit
B) African clawed frog
C) Guinea pig
D) Chicken
Now, the answer from last week’s TriviaRX: ✅ B) Advanced a catheter from his arm into his heart
Forssmann inserted a thin catheter through a vein in his arm into his own heart, then confirmed it with an X-ray. His work later helped establish heart catheterization as a major tool in cardiology.
That’s it for today.
As always, 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 week.
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).
