Machine leaning models to predict the risk of preterm birth

点击翻译按钮获取中文摘要

The Project will:-Cross-validate experimental data from placenta and fetal cell populations (cell mechanics, collagen, inflammatory biomarkers) with computational models to predict the risk of membrane failure to pollution.-Develop a biomarker database of individual characteristics linking clinical data from women (maternal and umbilical cord blood, amniotic fluid, intact placenta) with risk factors taken from women's demographics (gestational age, BMI, postcode location, work/home lifestyle, causes of PPROM).-Use experimental and clinical data to learn relationships and build machine learning tools that will automate analysis and predict women who are at greater risk of PPROM and preterm birth.The integrity of the fetal membranes that surrounds the baby in the womb during pregnancy is vital for normal development. Once the fetal membranes have ruptured or are damaged, they fail to heal leaving a defect until the end of pregnancy. This condition is called pre-term premature rupture of the fetal membranes (PPROM). Exposure to air pollution during pregnancy is associated with PPROM leading to preterm birth. Air pollution results mainly from the combustion of fossil fuels and from industrial emissions and has become a major global public health issue for all people in all age groups. In England, PPROM affects 1 in 9 pregnancies and in London, this increases to 1 in 7 annually. Unicef Executive Director warned the dangers of air pollution to unborn babies. Environmental harm during gestation, infancy and early childhood will have long term impacts across the lifetime. Currently, there are no clinical solutions to improve healing of the fetal membranes after they rupture.