Google AR-powered Microscope mixing Machine Learning and Augmented Reality has been developed for detecting cancer
Google AR-powered Microscope
Google AR-powered Microscope has been developed for detecting cancer. It’s basically a team effort by Google researchers who have come up with a Machine Learning (ML) and Augmented Reality (AR) powered microscope that helps in real-time detection for saving lives of the patients. During the annual meeting of the American Association for Cancer Research (AACR) in Chicago, Illinois on Monday, Google explained a prototype Augmented Reality Microscope (ARM) platform that helps in accelerating and democratising the adoption of deep learning tools for pathologists around the world.
It consists of a modified light microscope that makes real-time image analysis and presentation of the results of ML algorithms directly into the field of view. The ARM retrofits into existing light microscopes all around the world by using low-cost and easily available components. It doesn’t even need whole slide digital versions of the tissue being analysed. It has shown promising results in the field of pathology, dermatology, radiology, and ophthalmology.
Martin Stumpe, Technical Lead and Craig Mermel, Product Manager, Google Brain Team, wrote in a blog post, “In principle, the ARM can provide a wide variety of visual feedback, including text, arrows, contours, heatmaps or animations, and is capable of running many types of machine learning algorithms aimed at solving different problems such as object detection, quantification or classification.”
The post also said, “At Google, we have also published results showing that a convolutional neural network is able to detect breast cancer metastases in lymph nodes at a level of accuracy comparable to a trained pathologist.”
The new-age computational components and deep learning models, like those built on open source software “TensorFlow”, will allow a wide-range of pre-trained models for running on this platform. Google configured ARM for running 2 different cancer detection algorithms: one that helps in detecting breast cancer metastases in lymph node specimens and the second that helps in detecting prostate cancer in prostatectomy specimens.
According to Google, “While both cancer models were originally trained on images from a whole slide scanner with a significantly different optical configuration, the models performed remarkably well on the ARM with no additional re-training.”
Google added, “We believe that the ARM has potential for a large impact on global health, particularly for the diagnosis of infectious diseases, including tuberculosis and malaria, in the developing countries.”