Startup pitches AI blood test to diagnose postpartum depression

A healthcare startup in San Diego is pitching an AI-supported blood test to check for postpartum depression – a leading cause of maternal death – even before the symptoms appear. Researchers confirm that the method is working.

Postpartum depression is the most common complication of childbirth. While those with a mental illness are at increased risk, it occurs spontaneously in up to 15% of the population. That means it can happen to any mother.

Diagnosis and treatment is also spotty or even negligent, Dionysus Digital Health says. Now, the company has announced a new epigenetic blood test, Enlighten, that can be taken during pregnancy and that predicts future risk of postpartum depression.

According to the company, their researchers and academic partners have pinpointed a gene linking a person’s moods more closely to hormonal changes. Machine learning is involved.

AI is used to compare epigenetics – how genes are expressed – in the blood sample with benchmarks developed during years of research into pregnant people who did and did not develop postpartum depression.

Researchers at the Royal’s Institute of Mental Health Research and UVA Health – Dionysus’s academic partners – have already published peer-reviewed papers affirming their findings.

The new test could be of huge importance. One in seven mothers experience postpartum depression, and outcomes for the mother could be severe and include suicide risk. For infants, too, postpartum depression impairs healthy development.

Unfortunately, studies have demonstrated that a minority of pregnant patients in the United States who showed signs of mental disorders went on to receive treatment.

“A non-invasive private digital interface coupled with our groundbreaking AI acts as a first step to help you understand your risk. Anyone can take the PPD blood test,” Dionysus Digital Health says.

“Having a biological test showing you are at elevated risk to PPD will help you and your supporters to recognize symptoms if they develop. You’ll know you are more than ‘just tired’ and that it is time to get screened by your doctor so you can be treated.”

With potential benefits come risks, though. That’s because AI systems can be biased – just like the data sets they’re trained on.

In 2019, a study found that an algorithm recommending C-sections was wrongly flagging black women as high risk, and another algorithm consistently recommended less care for black patients.

Still, Dionysus Digital Health is hoping to eventually gain the US Food and Drug Administration’s approval and begin to sell the $250 test to the general public. The firm’s goal is also to have insurers cover the costs as well.

Startup pitches AI blood test to diagnose postpartum depression

More from Cybernews: