NIH awards Illinois $2.5M to improve IVF with advanced microscopy

A multidisciplinary collaboration at the University of Illinois Urbana-Champaign has received a $2.5 million award from the National Institutes of Health to develop technology for assessing embryo viability for in vitro fertilization.

The project will combine the microscopy and artificial intelligence expertise of Illinois Grainger Engineering electrical and computer engineering research assistant professor Haohua Tu and associate professor Alexander Gerhard Schwing, respectively, with the embryo production and developmental biology expertise of Illinois animal sciences professor Matthew Wheeler. Researchers will combine the “gentle,” label-free microscopy designed to minimize tissue damage developed by Tu with artificial intelligence developed by Schwing to analyze bovine embryos produced by Wheeler and predict viability rates when transferred in utero.

“In our work, we showed that microscopy based on nonlinear optical processes – a technology that is decades old – can be leveraged to give high-resolution images of biological tissue without damaging it,” Tu said. “For the technology’s entire existence, researchers have assumed that the nonlinear, or multi-photon, absorption damages the molecules it tries to image because of the high energies involved. Our group has shown this is not the case, and the risk of damage can be actively minimized.”

“The idea of the NIH grant is using this microscopy technique to improve in vitro fertilization, since it could allow embryo viability to be assessed before it is implanted,” Wheeler added. “We’re going to use it to study embryos implanted in cows – a good model species for human pregnancy – and see if it can predict which fetuses come to term. If successful, it would be very promising for human IVF.”

This work is made possible by Tu’s development of a nonlinear imaging protocol that achieves high-resolution images of biological tissue while minimizing the risk of tissue damage. He and his collaborators demonstrated that nonlinear, four-photon absorption and emission – the mechanism that makes high-quality imaging possible – is not responsible for molecular-level tissue damage as was previously assumed. The researchers found that low-energy linear absorption is the main mechanism for damage and showed that it is minimized by rapidly scanning light pulses across the sample instead of focusing light on one spot for extended lengths of time.

“We tested our system on several biological systems, including rat brain slices, nematode worms and even chicken breast tissue,” Tu said. “All of our studies support the hypothesis that linear absorption is the primary damage mechanism and can be mitigated by quickly scanning light across the entire sample. This way, the molecule has time to relax between absorption events.”

These results have been published in the journal Advanced Science.

This work formed the basis for the NIH proposal submitted by Tu and Wheeler, the latter of whom is also affiliated with the Department of Bioengineering in the Illinois Grainger Engineering. They believe that this scanning microscopy technique could allow for high-quality images of bovine embryos with minimal toxicity.

“The challenge with trying to image sperm, ova and embryos is that it’s hard to do without disturbing or damaging the system,” Wheeler said. “We took an interest in Professor Tu’s work both because it was shown to minimize damage to biological tissue and because it’s label free. There are no dyes or indicators to worry about disrupting fertilization and development.”

Wheeler and his colleagues will use Tu’s imaging system to monitor implanted bovine embryos as they develop in vitro. They believe that Schwing’s computer vision AI can be used to identify signals that indicate embryo viability, an otherwise difficult task.

“The only available tool for assessing embryo viability right now is the tincture of experience,” Wheeler said. “People spend years practicing it and developing an eye for it by looking at the embryos under an optical microscope. The idea is to combine the new microscopy with AI to detect the parameters that our eyes can’t.”

Wheeler’s animal sciences research group will fertilize 2,500 bovine embryos. Half will be scanned using Tu’s microscopy system, and the other half will be assessed by eye. They will then be shipped to California where they will be implanted in cows. The researchers hope that the combination of nonlinear microscopy and AI will outperform human examination when predicting which implants lead to birth.

Illinois Grainger Engineering Affiliations

Haohua Tu is an Illinois Grainger Engineering research assistant professor of electrical and computer engineering in the Department of Electical and Computer Engineering.

Alexander Gerhard Schwing is and Illinois Grainger Engineering associate professor of electrical and computer engineering in the Department of Electrical and Computer Engineering. He is also affiliated with the Siebel School of Computing and Data Science and the Coordinated Science Laboratory. He holds a W.J. "Jerry" Sanders III - Advanced Micro Devices, Inc. Faculty Fellow appointment.

Matthew B. Wheeler is an Illinois professor of biotechnology and developmental biology in the Department of Animal Sciences. He is also affiliated with the Department of Bioengineering, the Department of Veterinary Clinical Medicine, the Carl R. Woese Institute for Genomic Biology, the Beckman Institute for Advanced Science and Technology and the Carle Illinois College of Medicine.