Google makes microscope that can recognize cancer with artificial intelligence

During a conference in the United States, Google researchers described a prototype of a microscope using augmented reality and machine learning to help medical doctors recognize cancer cells in human tissue.

The so-called Augmented Reality Microscope is a slightly adapted regular microscope. The device works through the use of neural networks, which are first trained to recognize cancer cells in samples of human tissue. When a glass slide with human tissue is placed under the microscope, the image the physician sees through the viewfinder is loaded into a computer via a camera.
Google’s algorithms then try to detect cancer cells. If malignant cells are found, they are visualized when a researcher looks through the viewfinder of the microscope. According to Google, that happens with a delay of 100ms. The alerts can be applied in the form of text, arrows, contours, heat maps or animations.
The researchers tested the microscope’s capabilities in two types of cancer: breast cancer and prostate cancer. With these types, it appeared that Google’s microscope could detect fairly accurately whether there were cancer cells in the samples. According to Google, the prototype has the potential to detect other forms of cancer and other diseases, such as TB or malaria.
This process of analyzing the microscopic image and displaying where any cancer cells are located happens in real time . According to Google, the design for the microscope can be applied fairly easily to existing, light microscopes that can already be found in hospitals and clinics. The research is published under the title An augmented reality microscope for realtime automated detection of cancer .