Innovation born of U of A research licensed to Amahealth LLC to improve maternal health

TUCSON, Ariz. — Pregnancy is a time of great excitement juxtaposed with uncertainty or fear. One common challenge is that labor is unpredictable and the ‘due date’ is usually wrong, sometimes by a matter of weeks. Until now, expectant families and their care providers have been unable to predict with certainty when labor will start. Questions such as “When will I go into labor?” and “When should I go to the hospital?” may soon be easier to answer, thanks to technology developed through University of Arizona research. Arizona startup Amahealth LLC is developing a solution to provide these insights. 

The company’s name stems from the word “ama” or “amma,” which translates to “mother” in several languages. 

Image

Elise Erickson, associate professor of Physiology in the College of Medicine – Tucson and the BIO5 Institute and a certified nurse-midwife, and Shravan Aras, assistant director of the Center for Biomedical Informatics and Biostatistics and assistant research professor in the Mel & Enid Zuckerman College of Public Health, have been focused on translational research into the intersection of the physiological mechanisms that govern birth, the maternal variables that impact the onset of labor, and the clinical decisions that affect maternal health. Their work has given rise to an AI-powered, data-driven tool to forecast a personalized labor onset date and window for expectant families and their providers.

The team also includes Chief AI Officer Chinmai Basavaraj and Lead Scientist Azure Grant, both of whom are co-authors of the research behind the invention, and UX Research Lead Amy Goh. 

The team worked with Tech Launch Arizona, the university’s commercialization arm and a unit of the Office of Research and Partnerships, to protect the intellectual property behind the technology, strategize and launch Amahealth, and license the innovation to the company. 

“From the underlying research to the development of the technology and intellectual property to the launch of their startup, this team is the embodiment of the U of A mission of generating impact from the discoveries arising out of research,” said U of A Chief Innovation Officer and TLA Executive Director Doug Hockstad. “We look forward to continuing to support the team as they advance their research and their venture.”

Image

Erickson and Aras were inspired by research indicating that labor could be predicted by monitoring patterns in maternal vital signs. They then tested their ideas in human pregnancy, developing a deep neural network solution to provide an accurate labor forecast based on a pregnant individual’s personalized pregnancy progression. Currently, the timing of labor is unexpected until physical symptoms appear (e.g., contractions or water breaking), but to the mother’s body, labor is not unexpected. The body prepares for labor days and weeks in advance prior to birth. The team’s research showed that through patterns during pregnancy, their technology can detect the body’s natural process of preparing for labor by training an AI model using data from consumer wearables.

“I had the honor of providing care to families for nearly two decades as a midwife. Despite the many precision-based advances we have in medicine and health care, we have no answer to the universal question pregnant patients ask: ‘When will I go into labor?’.” Erickson says. “For many women and families, this information isn’t just a nice-to-know detail, because for them getting access to care during pregnancy requires driving a great distance to a birth center or hospital. According to a 2026 March of Dimes report, maternity care deserts are growing in Arizona and across the nation with 35% of counties having no maternal care facility. With a physiologically informed due date, we could offer the 5.8 million women living far from maternity care a better chance of getting pregnant patients the right care at the right time.” 

The original research underlying the innovation was supported by both state and federal funds. Most recently, the team has advanced their concept with experience gained through the NSF I-Corps program, which Tech Launch Arizona runs for U of A and regional community innovators and entrepreneurs. They also received funding from the Flinn Foundation, the Arizona Biomedical Research Centre and STTR funding from the National Institutes of Health. They also participated in the NIH Concept to Clinic: Commercializing Innovation (C3i) Accelerator, a 12-month program that provides technical development and business support to advance the translation of biomedical technologies. 

Image

According to Amahealth Chief Executive Officer Rachel Sprague, the company’s next milestones are to recruit a cohort of 1000 pregnant participants to be a part of a beta community to strengthen the algorithm’s accuracy and FDA regulatory submission. In exchange for contributing their data, participants will benefit from access to evidence-based education and localized resources, such as birth classes, support groups and postpartum resources.

“The unknown surrounding the timing of birth has rippling consequences across our healthcare system. But now, every mother can be given a personalized labor date,” said Sprague. “Our vision at Amahealth is to reduce uncertainty in pregnancy and postpartum, allowing more space for joy.” 

The femtech marketplace – one focused on software, diagnostics, products and services to improve health and well-being for women and those with female organs – is booming. According to PitchBook, global venture capital for such technologies is likely to reach $3 billion by the end of the decade. While dozens of direct-to-consumer pregnancy apps and care platforms exist, no products currently available directly address the uncertainty of labor onset by providing a real-time prediction of the duration of remaining pregnancy.