The DIU has awarded contracts to optics company Jenoptik to build the hardware, while Google develops the software. So far, 13 prototypes have been built, with one being tested for its ability to detect metastasized breast cancer in lymph nodes, achieving an accuracy level of about 94%. The technology is not yet ready for clinical use, but it shows promise. The AI-powered microscopes could be provided to the US military as early as this Autumn and may be sold to hospitals and clinics in the future for $90,000 to $100,000 each.
Key takeaways:
- Google has contributed software to the Augmented Reality Microscope (ARM), a device that uses AI to help pathologists identify cancerous cells in tissue samples more quickly.
- The ARM uses computer vision algorithms to highlight potentially problematic areas and can create heatmaps labeling cells as benign or cancerous.
- The Defense Innovation Unit (DIU) of the US Department of Defense is investing in the technology to aid pathologists in military hospitals, with the technology currently being used to detect breast, cervical, and prostate cancer, as well as cell duplication.
- Contracts have been awarded to optics company Jenoptik to build the hardware, and Google is developing the software. The technology is not yet ready for clinical use, but early experiments have shown it can classify cancerous and benign cells with about 94 percent accuracy.