Researchers at the University of Sydney have developed an ultra-compact AI chip using light for rapid and energy-efficient image computation, successfully classifying medical images including MRI scans.
Key Details
- 1Nano photonic AI chip performs neural computations using light, not electricity.
- 2Prototype was validated by classifying over 10,000 biomedical images, including MRI scans of breast, chest, and abdomen.
- 3Achieved classification accuracy of approximately 90-99% in experiments and simulations.
- 4Chip performs calculations on a picosecond timescale (trillionths of a second).
- 5The technology aims to enable faster, more energy-efficient AI processing with a minimal energy footprint.
- 6A patent has been filed and further work is planned to scale the technology.
Why It Matters
This innovation in AI hardware could drastically reduce power consumption and increase the speed of medical imaging AI, offering a sustainable solution as computational demands grow in clinical radiology. It may enable real-time analysis and more scalable deployment of imaging AI systems.
Source
EurekAlert
Related News
•EurekAlert
AI-Powered OCT Enables Rapid 'Optical Biopsy' for Early Endometrial Cancer Detection
A team at Washington University has developed a catheter-based 3D OCT system with AI to quickly and noninvasively detect early endometrial cancers.
•EurekAlert
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
•EurekAlert
KAIST, MIT, Microsoft Develop Efficient AI Image Upsampling for Robotics
KAIST, MIT, and Microsoft have created 'Upsample Anything,' a training-free AI method to restore high-resolution visual data from compressed images with up to 16x improved GPU memory efficiency.