Knowing exactly when an expectant mother conceived helps physicians treat life-threatening complications that come with early- and late-term childbirth. Yet parents often miss their estimates by weeks.

Physicists Abbas Ourmazd and Russell Fung are using a mathematical algorithm to estimate the time of conception with much greater accuracy. In some circumstances, the algorithm can reduce timing uncertainty by a factor of 300. If that holds true in this application, it could shrink the estimate to a range of days or hours. The project is funded by the Bill & Melinda Gates Foundation.

The computer algorithm works by extracting a weak arrow of time from noisy data with highly inaccurate time stamps. Think of it as restoring the initial sequence of a deck of cards after it has been heavily shuffled.

“There are some remnants of the original sequence in the shuffled deck,” says Fung, a senior scientist. “There’s a weak whisper of time, like a faint voice in a loud party.”

The algorithm, which has many different applications, was originally devised to make movies of ultrafast chemical bond-breaking in molecules, and Ourmazd and Fung published its details in the journal Nature.

“It is a privilege to use the algorithm we originally developed for fundamental science to help improve people’s health everywhere, particularly in developing nations,” says Ourmazd, a UWM distinguished professor of physics.

Neonatal health data will come from the foundation, which collects global statistics on the health of mothers and babies during, and for many years after, pregnancy.