A new paper in Biomedicines has found that a new deep learning-based method is effective in detecting both common and rare genetic disorders in fetal ultrasound scans, improving prenatal screening.
Despite recent advances in genetic testing, the majority of genetic diagnoses can only be performed after babies are born. This can cause delays in deciding on a treatment plan, and consequently can have negative impacts on the health of babies with genetic disorders. In an attempt to overcome this challenge, the authors of this paper developed Pgds-ResNet, a fully automated prenatal screening algorithm based on deep neural networks, which uses fetal ultrasound to detect abnormalities which indicate the presence of a variety of genetic diseases.
Following training on ultrasound images of fetuses affected by a range of different genetic disorders, including many rare diseases, Pgds-ResNet was found to perform better than other commonly used deep learning methods, and on par with senior sonographers. This indicates that it may be a useful tool to support obstetricians and sonographers in enhancing genetic disease screening during prenatal diagnosis. It could also be particularly useful in low- and middle-income countries, where resources are constrained and supportive tools are of particular need.
Source: Orphanet.
|
Subscripción via RSS |
·