MTAT magistritööd – Master's theses

Sirvi

Nüüd näidatakse 1 - 1 1

Gestatsioondiabeedi ja makrosoomia prognoosimine ning nende riskitegurite analüüs masinõppe meetoditega
(Tartu Ülikool, 2020) Pihu, Silvia; Laur, Sven, juhendaja; Rull, Kristiina, juhendaja; Tartu Ülikool. Loodus- ja täppisteaduste valdkond; Tartu Ülikool. Arvutiteaduse instituut
Large-for-gestational-age (LGA) may cause problems for both baby and mother during delivery, therefore the best solution is to predict and avoid it (by diet, cure of GDM, etc.) or at least use planned Caesarian section. Gestational diabetes (GDM) is known as a major risk factor for too large baby. Different machine learning algorithms were used to predict GDM and LGA on Estonian pregnancies and newborn data from 2012 to 2018 (4787 cases), using their risk factors. The best results were obtained by random forest method (AUC for GDM 0.96 and for LGA 0,92). The major risk factors for LGA occurred to be GDM and its correct diagnosing, the body mass index of the mother (before pregnancy), having large baby in previous pregnancy, the age of mother and the blood sugar level registered at the beginning of pregnancy.