Enhancing the workhorse: Artificial intelligence, hardware innovations boost confocal microscope's performance



Mouse esophageal tissue slab was imaged in triple-view sim mode. Yicong Wu and Xiaofei Han are authors of Nature.

Since Marvin Minsky patented the principle of confocal microscopy in 1957, it has become the workhorse standard in life science laboratories. Confocal microscopes are not perfect. It can take quite a while to do a full scans of a biological sample, exposing it to light dosages that can be toxic.

A collaboration at the Marine Biological Laboratory has invented a "kitchen sink" confocal platform that borrows solutions from other high-powered images, and adds a unifying thread of "Deep Learning" artificial intelligence. Their report on the technology, called Multiview Confocal Super-Resolution Microscopy, is published online this week in Nature.

Many labs have confocals, and if they can use artificial intelligence to improve their performance, then they don't have to invest in a new microscope. Hari Shroff is a senior author and MBL Fellow of the National Institute of Biomedical Imaging and Bioengineering.

The new confocal platform uses three objective lens to image a wide variety of sample sizes. Multiple specimen views are rapidly captured, registered and fused to yield reconstructions. The platform also has innovative scanning heads for the three lenses that allow illumination to be added to the microscope base.

The human osteosarcoma cell has been fixed and expanded. The triple-view deconvolved result is compared with the single view C.

The team added "super-resolution" capacity to the platform by adapting techniques from structured illumination microscopy.

The hardware summit that gets climbed in this platform is the multiple lenses around the sample, and then the super-resolution trick, which takes a combination of hardware and computation to achieve. It's a tour de force, but it's a phototoxic recipe. Patrick La Rivire of the University of Chicago said there was a lot of light being delivered to the sample.

Reducing the light coming from the microscope's laser is one way to address phototoxicity. Background graininess can obscure fine details of the object you want to image, and that's what you start having problems with. Artificial intelligence comes in here.

Whole, fixed, stained adult worm. The first part of the movie compares reconstructions with different types of confocals. The second movie has a reconstruction with 2136 segments. Yicong Wu and Xiaofei Han are authors of Nature.

The team trained a Deep Learning computer model to distinguish between better and poorer images with a low signal-to-noise ratio. "Eventually the network could predict the higher SNR images, even given a fairly low SNR input," Shroff said.

La Rivire said that Deep Learning allows you to take a hardware summit as the gold standard for resolution and then train a neural network to achieve similar results with much lower SNR data, many fewer acquisitions, and so much less light dose to the sample.

The platform's capabilities were demonstrated on more than 20 different fixed and live samples, targeting structures that ranged from less than 100 nanometers to a millimeter in size. There were distributions of genes in single cells, myoblasts in the wing of the Drosophila melancholia, and brain tissues. Potential applications for human tissue in histology and pathology labs are seen by them.

Shroff, La Rivire and co-author Daniel Coln-Ramos of Yale School of Medicine have been collaborating at MBL for nearly a decade to develop technologies with higher speed, resolution and duration. Applied Scientific Instrumentation is a company that they worked with both at MBL and the National Institutes of Health.

The first author on the paper built the new confocal platform. The pilot version of the new course,DL@MBL: Deep Learning for Microscopy Image Analysis, was taught by Wu. La Rivire is a faculty member.

Shroff said that the course taught Yicong how to learn Deep Learning methods in 4 days and quickly innovate with them, so they can now apply them in their lab. That's a short feedback scheme. It was great that MBLcatalyzed it.

More information about Yicong Wu, Multiview confocal super-resolution microscopy. Nature.com has an article titled "S41586-021-04110-0".

Nature journal information.

Enhancing the workhorse: Artificial intelligence, hardware innovations boost confocal microscope's performance was retrieved on November 24, 2021.

The document is copyrighted. Any fair dealing for the purpose of private study or research cannot be reproduced without written permission. The content is not intended to be used for anything other than information purposes.