AI can ‘follow’ growing babies: it can predict preterm birth as early as 31 weeks

According to the Centers for Disease Control and Prevention (CDC), about 10% of all babies born in the US in 2021 were preterm, meaning they were born before 37 weeks of gestation.

Preterm births also account for about 16% of infant mortality.

Now researchers at Washington University in St. Louis. Louis, Missouri, are looking to improve those odds by use of artificial intelligence.

They developed a deep learning model that can predict preterm labor by analyzing the electrical activity in a woman’s uterus during pregnancy and then tested the model in a study published in the medical journal PLOS One.

AI TOOL HELP DOCTORS UNDERSTAND CHAOTIC PATIENT DATA AND IDENTIFY DISEASES: ‘MORE MEANING’ INTERACTIONS

“The key finding is that it is possible to get data as early as week 31 and predict preterm birth before week 37,” which surprised the researchers, Arie Nehorai, Ph.D., professor of electrical engineering at University of Washington at LouisFox News Digital reported.

woman on ultrasound

Researchers at Washington University in St. Louis. Louis, Missouri developed a deep learning model (not shown) that can predict preterm birth by analyzing the electrical activity in a woman’s uterus during pregnancy. (iStock)

AI/deep learning automatically learned the most informative signs from the data that are relevant to the prediction of preterm birth,” he added.

In addition, the results show that preterm birth is an abnormal physiological condition and not just pregnancy that ended prematurelyNehorai said.

During the study, researchers performed electrohysterograms (EGGs), which use electrodes on the abdomen to record electrical activity in the uterus.

They took records of these electrical currents from 159 pregnant women who were at least 26 weeks pregnant and “trained” an AI model on the data.

RUPTURED BRAIN ANEURYSM GETS INFLUENCE ON SOCIAL NETWORKS IN MEDICATED COMA AFTER EMERGENCY COP

They combined this data with medical information such as the woman’s age and weight, the weight of the fetus, and any bleeding that occurred during the first or second trimester.

Nearly 19% of the women in the study gave birth prematurely. Theoretically, these women’s data can be used as a guideline for predicting preterm birth.

Premature baby

About 10% of all babies born in the US in 2021 were preterm, according to the CDC, meaning they were born before 37 weeks of gestation. (iStock)

“The advantage of our approach is that it is inexpensive to build,” Nehorai said of the new study. “Our model was effective in predicting shorter EGG recordings, which could make the model easier to use, more cost-effective in the clinical setting, and possibly suitable for home use.”

Looking ahead, the researchers believe that this method should be adopted by hospitals and obstetricians as part of routine check-ups for pregnant women. This will allow pregnant women to seek medical care and make lifestyle changes to protect their baby’s health as needed.

“Our work furthers the goal of using EGG measuring devices to accurately predict preterm birth.”

“For this purpose, it is necessary to create a device designed to implement our method,” Nekhoray said.

It’s hard to say how long it will be before such a test becomes widely available, the researchers said.

“There are already some devices on the market to measure EGG, but it has been difficult to predict preterm birth based on EGG data,” said Uri Goldstein, Ph.D. in the Department of Biomedical Engineering, working under Professor Nehorai in Washington. University.

NEW CANCER AI CHATBOT PROVIDES PATIENTS AND FAMILIES 24/7 SUPPORT: A “SENSITIVE APPROACH”

“Our work furthers the goal of using EGG measuring devices to accurately predict preterm labor,” he told Fox News Digital.

EGG measurements typically take 30 to 60 minutes, with additional time required to place the device on the mother’s abdomen, Goldstein noted.

“We have shown that predictions can be made from shorter EGG measurements, less than five minutes long, without unduly degrading prediction accuracy,” he told Fox News Digital. “This finding is significant because the long duration of EGG measurement is an important limitation for its use in clinical settings.”

The promise of deep learning, but with caveats

Dr. Susie Lipinski, Board Certified Ob/Gynecologist at Pediatrix Medical Group in Denver, Coloradodid not participate in the study, but shared her opinion on whether deep learning technology can help solve the problem of preterm birth in the United States.

“Being able to predict who is at risk before they go into labor would be very helpful,” Lipinski told Fox News Digital. “Using a deep learning model seems promising, however, this study included a relatively small number of patients, so it is not possible to determine how applicable this is to a larger population.”

“Being able to predict who is at risk before they go into labor would be very helpful,” one obstetrician told Fox News Digital. (iStock)

Previous research using AI did not show high reliability, so more studies and more patients will be required before we start using this method, ”she added.

The doctor pointed out that another potential limitation is that very few places use EGG measurements.

“The standard in most hospitals and offices is to use a current dynamometer that measures pressure rather than electricity,” she explained.

OZEMPIC, WEGOVY AND THE RISK OF PREGNANCY: WHAT YOU NEED TO KNOW ABOUT THE PROBLEM

If EGG becomes a way to assess preterm birth, Lipinski said, hospitals, maternity clinics and offices will have to procure new equipment, which could delay rollout in low-resource areas such as rural and central cities.

“The higher preterm birth rate in this study than the national average also raises questions about its applicability,” she told Fox News Digital. “Demographics of the patient population were not provided, so it’s impossible to see how they reflect the population of the entire country.”

“Being able to predict who is at risk before they go into labor would be very helpful.”

There is also the potential for false positives, Lipinski said.

“While this method predicts better than our current methods, there are still many patients who will be identified as being at risk who may not have preterm births,” she said. “This false positive result will cause a greater burden on the patient as well as increased utilization. health resources.”

If and when this becomes the new standard of care, Lipinski said, there will need to be better treatments for preterm birth.

“Our problems with preterm birth are twofold: we have a poor prognosis, but also poor prevention options after 26 weeks,” she added.

Researchers share key study limitations

According to Goldstein, the study has two major limitations.

“First, we developed our work using about 160 samples from two publicly available datasets,” he said. “While this amount of data was sufficient for our initial study, a much larger data set would have been required to develop and validate a medical product.”

Pregnant woman at the doctor

According to the lead researcher, the results of the study show that preterm birth is an abnormal physiological condition and not just a pregnancy that ended prematurely. (iStock)

The second limitation is due to the nature of deep learning, which can produce accurate results but is usually difficult to interpret, Goldstein says.

“In other words, it is difficult to understand how the algorithm makes predictions,” he explained.

CLICK HERE TO SUBSCRIBE TO OUR HEALTH INFORMATION

Discussing the results in a medical journal, the authors noted that “although machine learning algorithms can help improve healthcare, and many studies are making headway in this area, important challenges remain.”

“To develop and validate a medical product, a much larger set of data will be needed.”

Among these problems, the researchers wrote, it can be difficult to determine the reasons behind the algorithm’s predictions.

CLICK HERE TO GET THE FOX NEWS APP

“In our case, while our predictions may affect the management of the pregnancy, our predictions will need to be supplemented by additional medical assessments to determine which treatments are more likely to reduce the risk of preterm birth and improve outcomes,” the researchers also said.