New Scientist - 08.02.2020

16 | New Scientist | 8 February 2020

A NEW method for manipulating the

quantum state of particles has been

used to chill a tiny glass bead into

its coldest possible quantum state.

Once you get down to extremely

small scales, heat and motion are

interchangeable: the more a particle

is moving, the hotter it is. So to cool

down a small particle, you have to

stop it moving. Because the rules

of quantum mechanics mean you

can never know exactly how fast

a particle is moving, there is a limit

to how cold a particle can get. When

a particle is at that limit, we say

it has reached its ground state.

Markus Aspelmeyer at the

University of Vienna in Austria and

his colleagues used a single laser

to cool a 150-nanometre-wide

glass particle into its ground state

(Science, doi.org/dkvw).

The laser levitated the particle

through an effect called optical

trapping, in which the light

interacted with the particle to hold

it in place. Mirrors either side of the

particle caused the light to overlap

and interfere with itself.

This interfering light could

only exist in certain frequencies,

according to quantum mechanics.

That allowed the researchers to

select the precise frequencies of

light that hit the particle. As the

particle vibrated back and forth,

some frequencies of light would

have sped it up by imparting small

amounts of energy, while others

would have slowed it down by

removing energy.

Allowing only the frequencies

that slowed the particle meant

it kept cooling until it reached its

ground state. This occurred at a

temperature of 0.000012 kelvin

(about -273°C), just a fraction

above absolute zero, the minimum

temperature possible. ❚

ARTIFICIAL intelligence is helping

to select embryos for IVF that

have a higher chance of resulting

in a successful pregnancy.

The AI algorithm, called Ivy,

analyses time-lapse videos of

embryos as they are incubated

after being fertilised, and

identifies which ones have the

highest likelihood of success.

It was developed by Harrison.ai,

a tech firm in Sydney, and has been

used for several thousand women

undergoing IVF in Australia.

People who undergo IVF using

Ivy are informed about the

algorithm and consent to its use.

Ivy was trained on more than

10,000 videos of embryos growing

inside an incubator for five days,

in combination with data about

which embryos resulted in

pregnancy. The model is more

accurate than trained specialists

at predicting the likelihood that

an embryo will go on to develop

a fetal heartbeat at or beyond

seven weeks of gestation.

In a study of its predictive

ability, Ivy scored 0.93 on a

measure known as AUC, where

a score of 1 indicates a model

with predictions that are 100 per

cent correct. Trained human

embryologists have previously

scored an AUC of around 0.74.

“The results look very

promising,” says Iman

Hajirasouliha at Cornell University

in New York. He and his colleagues

have developed a similar AI, Stork.

What these AIs are identifying

in the videos as a marker of

success isn’t entirely clear.

“It’s still a bit of a black box,” says

Aengus Tran, CEO of Harrison.ai.

“We could speculate that it must

have learned very similar features

to what embryologists are looking

at,” says Tran. These might include

the shape of cells or how long it

takes to go from two to eight cells.

Ivy may also have picked up

on features that human specialists

don’t look at, says Tran. While

embryologists look only at video

snippets of key events of an

embryo’s development, Ivy

is able to rapidly analyse the

entirety of five days of footage.

Ivy is now routinely used as a

tool in embryo selection by IVF

Australia, which provides about

20 per cent of the country’s IVF

services. “Because we think

it’s a better predictor than

the embryologists, we are

choosing the embryos with

Ivy most of the time,” says Peter

Illingworth at IVF Australia.

A clinical trial will begin next

month to determine whether

people undergoing IVF whose

embryos are selected by Ivy

become pregnant more quickly

than those whose embryos are

selected by human specialists.

In an average IVF cycle, a parent

starts with 10 eggs that result in

three potentially viable fertilised

embryos after five days, says

Illingworth. One is implanted,

and the rest are frozen and stored.

“Ultimately, what you really

want is one IVF cycle resulting

in one embryo being transferred

and one healthy baby,” says Tran.

“What we’re not doing is

using Ivy to decide which

embryos we will freeze and

keep,” says Illingworth. “Ivy

will never make the difference

between a woman having a

baby or not having a baby.” ❚

Quantum physics

Eggs undergoing IVF have

different probabilities of

developing into embryos

CHRISTOPH BURGSTEDT/SCIENCE PHOTO LIBRARY

News

-273°C

Temperature the nanoparticle

was cooled to

Nanoparticle

quantum chilled to

near absolute zero

Leah Crane

Machine learning

Donna Lu

AI attempts to pick embryos with

the best chance of IVF success