top of page
Latest Innovations
Adoption before approval
Illustration by Alicia Tatone. Source: Getty 80% of physicians are already using AI at work. Most of the generative AI tools they use have not been approved by the FDA. A Science study found ChatGPT more accurate at diagnosis than hundreds of practicing doctors. Even the Harvard physician who led the research expressed concern about what the results might mean. Medicine likes to think it moves at the speed of evidence. In reality, it often moves at the speed of utility. AI is
The robot learns about the hospital before entering it
Source: NVIDIA Developer Most hospital automation projects never reach clinical deployment because validating robots in real healthcare environments remains difficult. NVIDIA Rheo uses digital twins of real hospitals to train and test robots before deployment. Same layouts. Same lighting. Same workflows. Simulation is becoming a critical layer of the hospital robotics stack. Read more: NVIDIA developer
AI triages breast cancer
Source: Noah Berger Women with abnormal mammograms often wait weeks to find out if they have cancer. UCSF used an AI model called Mirai, trained on hundreds of thousands of mammograms, to identify the 12.7% of patients most likely to have cancer. Those patients went from screening to evaluation to biopsy in a single day. Wait time for a diagnostic evaluation dropped from several weeks to about an hour. The wait time for a biopsy dropped from two months to under 10 days. The A
All Innovations
Adoption before approval
Illustration by Alicia Tatone. Source: Getty 80% of physicians are already using AI at work. Most of the generative AI tools they use have not been approved by the FDA. A Science study found ChatGPT more accurate at diagnosis than hundreds of practicing doctors. Even the Harvard physician who led the research expressed concern about what the results might mean. Medicine likes to think it moves at the speed of evidence. In reality, it often moves at the speed of utility. AI is
The robot learns about the hospital before entering it
Source: NVIDIA Developer Most hospital automation projects never reach clinical deployment because validating robots in real healthcare environments remains difficult. NVIDIA Rheo uses digital twins of real hospitals to train and test robots before deployment. Same layouts. Same lighting. Same workflows. Simulation is becoming a critical layer of the hospital robotics stack. Read more: NVIDIA developer
AI triages breast cancer
Source: Noah Berger Women with abnormal mammograms often wait weeks to find out if they have cancer. UCSF used an AI model called Mirai, trained on hundreds of thousands of mammograms, to identify the 12.7% of patients most likely to have cancer. Those patients went from screening to evaluation to biopsy in a single day. Wait time for a diagnostic evaluation dropped from several weeks to about an hour. The wait time for a biopsy dropped from two months to under 10 days. The A
A tumor removed through the eyelid
Source: The Irish Times Surgeons in Dublin removed a brain tumor through a patient's eyelid. A 2cm incision at the eye socket replaces a 15 to 20cm craniotomy. The eye socket becomes the corridor to the brain. Less trauma. Fewer complications. Patients go home in two or three days instead of two weeks. This is what surgical innovation looks like. Small changes in approach. Massive changes in outcome. Read more: The Irish Times
The hand that teaches the robot
Surgeons spend years developing hand dexterity. Until now, that expertise has been impossible to capture. Researchers at MIT built an ultrasound wristband that lets AI infer finger movements in real time by imaging the tendons inside the wrist. If surgical skill can be converted into data, it could eventually be stored, shared, and used to train the next generation of surgical robots. The future of autonomous surgery may begin long before a robot enters the operating room. So
A beating heart you can program
Every year, surgeons operate on diseased heart valves without being able to rehearse on a model that behaves like the real thing. UNSW Sydney just built one. A soft robotic heart with artificial muscles, silicone chambers, and working valves that leak, prolapse, and fail exactly like human disease. Powered by hydraulic pressure. Compatible with ultrasound imaging. They also tested a surgical catheter inside it. The goal is to create patient-specific models that enable surgeon
The Documentation Burden
Source: Stanford Medicine A discharge summary takes 15 to 30 minutes to write. Stanford built an AI agent that generates one automatically every morning for each patient. Physicians used it. Burnout scores dropped. Hallucinations were rare at 2%. The tool was not designed to replace clinical judgment. It was designed to eliminate the part of the job that exhausts physicians before they even begin. Read more: Stanford Medicine
bottom of page