Mirvie
Draw one tube of blood to predict preeclampsia risk, warning months earlier with RNA plus machine learning
Mirvie uses a very different route: analyzing cell-free RNA in a pregnant woman's blood, combined with machine-learning interpretation, to predict the risk of preeclampsia, preterm birth, and fetal growth restriction months before symptoms appear.
Features and application scenarios
Preeclampsia is one of the most dangerous complications of pregnancy, and the trouble is that current risk assessment mainly relies on medical history and clinical factors like blood pressure — for pregnant women without obvious high-risk factors, it's often not discovered until symptoms appear. Mirvie's RNA platform combines transcriptome analysis with AI: about 22,000 cell-free RNA transcripts are analyzed per subject, equivalent to directly reading real-time signals of placental and pregnancy biology rather than inferring from indirect clinical indicators. Supporting this method is a study of rare scale — covering nearly 11,000 representative pregnancies across the US, totaling about 200 million data points, with the results published in Nature Communications. The validation results announced in April 2025 showed that for pregnant women over 35 with no pre-existing high-risk factors, this blood test could identify 91% of cases that subsequently developed early-onset preeclampsia — that is, nine of every ten caught. The commercialized product is named Encompass, assessing placental health and pregnancy biological state by analyzing cell-free RNA. The research team has also published results on using the same platform to find molecular signals of severe fetal growth restriction.
Suited to OB-GYN clinical teams and pregnancy-care systems; for Taiwanese readers, this is an important case for understanding "how liquid biopsy and AI enter routine prenatal care," with actual testing currently mainly in the US market.
Main features
- Analyzes cell-free RNA in blood to assess pregnancy risk
- Analyzes about 22,000 RNA transcripts per subject
- Combines transcriptome analysis with machine-learning interpretation
- Encompass: preeclampsia-risk-assessment test product
- Study covered nearly 11,000 pregnancies, about 200 million data points
- Can provide risk prediction months before symptoms appear
- The same platform extends to preterm-birth and fetal-growth-restriction research
Common uses
- Preeclampsia-risk assessment for older pregnant women
- Supplementary screening for pregnant women with no obvious high-risk factors
- Risk-stratified scheduling of prenatal-visit frequency and monitoring intensity
- Early identification of preterm-birth risk
- Molecular-level research on fetal growth restriction
Key Features
- Analyzes cell-free RNA in blood to assess pregnancy risk
- Analyzes about 22,000 RNA transcripts per subject
- Combines transcriptome analysis with machine-learning interpretation
- Encompass: preeclampsia-risk-assessment test product
- Study covered nearly 11,000 pregnancies, about 200 million data points
- Can provide risk prediction months before symptoms appear
- The same platform extends to preterm-birth and fetal-growth-restriction research
Pros
- Directly reads biological signals rather than relying on indirect clinical risk factors
- Large study scale with results published in a peer-reviewed journal
- Still has a 91% detection rate for the group with no pre-existing high-risk factors
- Only a blood draw needed, with low burden on the pregnant woman
Cons
- Currently mainly in the US market; no widespread access channel in Taiwan yet
- A high-end self-pay test with undisclosed price
- Predicts risk rather than confirming a diagnosis, still requiring clinical follow-up and management
- Validation data concentrated on specific group conditions; performance in other groups needs more research
Use Cases
- Preeclampsia-risk assessment for older pregnant women
- Supplementary screening for pregnant women with no obvious high-risk factors
- Risk-stratified scheduling of prenatal-visit frequency and monitoring intensity
- Early identification of preterm-birth risk
- Molecular-level research on fetal growth restriction
Editor's Note
This is one of the most convincing medical-AI cases I've seen this year — not because of what model it used but because it first did a nearly-11,000-case study published in Nature Communications, then talked about the product. Too much medical AI does it backwards, issuing press releases first and adding evidence later. Taiwanese readers can't use it for now, but this 'evidence before product' order is well worth local teams' reference. (This article is an information roundup, not medical advice; discuss any pregnancy testing and management with your OB-GYN.)
FAQ
If the test shows high risk, will preeclampsia definitely occur?
No. It gives a risk probability rather than a confirmed diagnosis; its use is to let the medical team increase monitoring intensity and intervene earlier. Conversely, low risk doesn't equal zero risk — pregnancy care should still proceed routinely, and no test can replace continuous clinical follow-up.
Can pregnant women in Taiwan get this test?
It's currently mainly offered in the US market, with no widespread access channel in Taiwan yet, nor is it included in routine prenatal care. If needed, you should first discuss Taiwan's existing preeclampsia-screening methods with an OB-GYN rather than trying to send a sample yourself.
How is cell-free RNA testing different from ordinary prenatal genetic testing (NIPT)?
NIPT looks at cell-free DNA, mainly screening for chromosomal abnormalities; Mirvie looks at cell-free RNA, reflecting genes' current expression activity — the 'present continuous tense' of the placenta and pregnancy. The two have different purposes and don't replace each other.
Related AI Tools
Motobook
Taiwan’s first used motorcycle transparent pricing platform featuring over 37,000 listings and 812 models.
Carbook
Taiwan Used Car Price Registry – Real Market Values, Inventory, and Depreciation at a Glance
實價雷達 HouseTW
A free Taiwanese real estate platform overlaying 3.49 million official transaction records with soil liquefaction, fault lines, and flood risk maps.
Perplexity
AI search and answer tool with source citations.
Incentivio
Customer engagement platform for multi-unit restaurants with AI-powered churn prediction.
HeyDonto
A semantic integration layer for healthcare data that aligns siloed dental and oncology data into actionable signals.