Fast, Highly Accurate AI Diagnoses

New AI Tech Spots Pneumonia and Eye Disease

Scientists at the University of California San Diego have developed an artificial intelligence system that accelerates treatment of pneumonia and retinal disease.  The diagnoses take 30 seconds and the accuracy is close to 100%.

Transfer Learning on 200,000 Patient Scans
It's a computational tool using AI and machine learning techniques.  The technique is called transfer learning in which knowledge gained in solving one problem is stored in the computer and applied to a different but similar problem.  The AI system reviewed 200,000 patient scans.

Highly Accurate Results
This new AI system, first reported in the journal Cell, has spot-on accuracy.  The accuracy rate for eye disease is 95% and for pneumonia, differentiating between viral and bacterial pneumonia, is 90%.  The turnaround time is 30 seconds.

Future of AI in Medicine
The UC San Diego researchers say the future of better, cost effective patient care is through more data, more computational power and more experience on the part of the professionals using emerging AI systems.