Search
Menu

Artificial Intelligence Brings Specificity to Label-Free Imaging for Assisted Reproduction

Facebook X LinkedIn Email
URBANA, Ill., Aug. 4, 2020 — As part of a project to develop specific traits in cattle that would make them resistant to heat and diseases in tropical regions, researchers at the University of Illinois at Urbana-Champaign developed a phase-imaging technique to determine which sperm samples would work best for in vitro fertilization. This technique is being further developed for humans.

The approach combines label-free imaging and artificial intelligence to obtain nondestructive markers for reproductive outcomes. The phase-imaging system shows nanoscale morphological details from unlabeled cells, while deep learning provides a structural specificity map for segmenting with high accuracy the head, midpiece, and tail of the sperm.

“We knew from the fertilization experiments which sperm samples worked. We used our imaging technique to understand what parameters were important for success,” researcher Mikhail Kandel said.

Phase imaging with computational specificity applies artificial intelligence to label-free spatial light interference microscopy data to map subcellular compartments. Courtesy of the Beckman Institute for Advanced Science and Technology.
Phase imaging with computational specificity applies artificial intelligence to label-free spatial light interference microscopy data to map subcellular compartments. Courtesy of the Beckman Institute for Advanced Science and Technology.

To introduce specificity to the label-free images, the researchers trained a deep convolutional neural network to perform semantic segmentation on the quantitative phase maps. This approach, a form of phase imaging with computational specificity (PICS), allowed the researchers to efficiently analyze thousands of sperm cells and identify correlations between dry-mass content and artificial-reproduction outcomes.


The researchers measured the dry-mass content of each component and found that the dry-mass ratios represented intrinsic markers with predictive power. “We saw that the relationship between the size of the head and the tail of the sperm is an important parameter for fertility,” Kandel said.

Artificial intelligence was also used to automate the process of analyzing the sperm cells and speed the phase-imaging process, researcher Yuchen He said. The team hopes to improve the speed of the technique further for future analyses.

“For many years, we have developed various techniques for label-free imaging knowing that we had to give away molecular specificity,” professor Gabriel Popescu said. “However, our newly developed phase imaging with computational specificity brings back the molecular specificity via AI, which is harmless and works on live cells.

“The applications are limitless, but one that truly benefits from absence of chemical stains is the assisted reproduction, as described in this collaborative study.”

The research was published in the Proceedings of the National Academy of Sciences (www.doi.org/10.1073/pnas.2001754117).

Published: August 2020
Glossary
machine learning
Machine learning (ML) is a subset of artificial intelligence (AI) that focuses on the development of algorithms and statistical models that enable computers to improve their performance on a specific task through experience or training. Instead of being explicitly programmed to perform a task, a machine learning system learns from data and examples. The primary goal of machine learning is to develop models that can generalize patterns from data and make predictions or decisions without being...
artificial intelligence
The ability of a machine to perform certain complex functions normally associated with human intelligence, such as judgment, pattern recognition, understanding, learning, planning, and problem solving.
Research & TechnologyeducationAmericasUniversity of Illinois at Urbana-ChampaignImagingphase imaginglabel-free imagingneural networksconvolutional neural networksTest & MeasurementMicroscopyLight SourcesBiophotonicsassisted reproductionmachine learningartificial intelligencespermmale infertility

We use cookies to improve user experience and analyze our website traffic as stated in our Privacy Policy. By using this website, you agree to the use of cookies unless you have disabled them.