Yesil Science / Blog
Blog
Every article we have published: research summaries, product thinking and developments in health AI.
Latest articles

15 NHS Sites Connect Digital Pathology Network
Connecting 15 hospital sites to a single digital pathology platform proves infrastructure can scale, but sharing slides across trusts does not solve...

Fetal monitor signal loss warns of asphyxia risk
Missing fetal monitoring data is not just technical noise; high signal dropout during internal heart tracking strongly correlates with severe birth asphyxia.

Federal Health Agencies Join TEFCA Network
Clinicians treating millions of veterans and service members can now retrieve civilian medical records through a single national framework.

Machine Learning Predicts Chiari Syrinx Shunt Need
Delaying decompression surgery for Chiari malformation directly raises the risk that a patient will need a second invasive brain operation later.

AI Models Flunk Complex Pharmacy Exam Questions
High scores on basic medical recall hide critical AI failures in complex clinical reasoning and visual data.

AI Finds Cancer Mutations Directly From Routine Slides
Routine tissue slides can now alert oncologists to targeted treatment targets before tissue runs out or sequencing delays care.

Patients Trust AI More Than Doctors Expect
Clinicians are severely underestimating how ready patients are for artificial intelligence in the clinic.

ER Doctors Would Act on Fake AI Abstracts
Polished writing from AI models tricks acute care physicians into trusting fabricated scientific claims.

Over 170,000 Emergency Visits Diverted by Triage Service
Free nurse-led phone triage is quietly reshaping how patients access acute care before reaching emergency departments.

DeepSeek Matches GPT Accuracy But Fails Safety Tests
Open-source AI models can match proprietary giants in diagnostic skill, but their safety guardrails collapse under deliberate pressure.

Nurse Practitioner Network Secures $15M for Medicare ACO
Independent nurse practitioners are securing capital to run value-based care networks without physician ownership.

Fine-tuned LLM accelerates trial bias reviews
Pairing a fine-tuned language model with clinician reviewers cuts manual overhead in clinical trial appraisals without sacrificing judgment accuracy.
Also on LinkedIn
AI Health Hub posts the same research summaries on LinkedIn every day. More than 3,000 people follow it there.