Researchers at Carnegie Mellon University (CMU) have developed a machine learning technique that
For example, one of the signs of future complications withpregnancy is damaged blood vessels, vasculopathy. Their presence suggests that complications may occur with 2-8% of pregnancies. In extreme cases, they can be fatal to both mother and baby.
If this damage is detected at an early stage,they can be treated before symptoms appear. However, because examinations are very time-consuming and require highly specialized skills, they are rarely done.
A new test will determine its safety in a few minutes and with just a drop of water.
The team trained their algorithm based onimages of placenta samples. It first detects all the blood vessels in the image and then determines whether each individual vessel is healthy. The algorithm also evaluates various features of the pregnancy, such as the duration and any conditions the mother is experiencing. If the system detects any abnormalities, it assesses the risk of complications.
When testing, the algorithm classifieddamage is more accurate than professional pathologists. However, researchers do not expect the system to replace living specialists. Instead, the system will select a sample for doctors to look at. The researchers note that this will reduce the cost of surveys.
See also:
- The USA approved the smallest nuclear reactor in the world
- Simulation showed what dark matter would look like
- Scientists have found out why children are the most dangerous carriers of COVID-19