Beckman Institute for Advanced Science and Technology researchers developed a Covid-19 test that is rapid, accurate, and cost-effective by combining microscopy and artificial intelligence. As the pandemic spreads around the globe, screening procedures must stay flexible and intelligent. The initial step taken by the research team was to discover innovation potential. While there were various Covid-19 testing systems available, none of them used a label-free optical approach. Even with a microscope, relying just on sight is nearly impossible due to the small size of a single particle.
While electron microscopy can be used to image the structure of a particle, substantial preparation is required to ensure that the sample is visible. Therefore, the researchers chose to experiment with a technology developed at Beckham that is commonly used to visualize cells.
The image produced by an electron microscope is clear, but it necessitates substantial sample preparation. It’s like staring at something without your glasses on when you use SLIM for viral imaging. Because the viruses are smaller than the diffraction limit, the image is hazy. However, because of SLIM’s exceptional sensitivity, we can identify and distinguish between different types of viruses.Based on SLIM data and artificial intelligence, the research team discovered a novel approach to detect Covid-19.
An advanced deep neural network can recognize even the most blurry photos with the proper training. A stained Covid-19 particle emitting fluorescence and a phase image acquired with a fluorescence-SLIM multimodal microscope were used to train the AI program. The AI had been programmed to recognize them as being the same. After introducing the AI to recognize , the machine learning technology was taught to distinguish from other viruses and particles was distinguished from other viral infections such as H1N1, adenovirus, and Zika virus by the AI. The preclinical trial was a success, with a Covid-19 detection rate of 96 percent.