22 June 2026 : Clinical Research
Development and Validation of Machine-Learning-Based Prediction Models for Thyroid Diseases During Pregnancy
Guang Yang BCDEF 1, Yi Gao BCEF 1, Pengfei Liu BC 1, Jingwen Jiang BC 1, Hui Qiao ADEG 1*, Weixuan Sheng
ACDE 1
DOI: 10.12659/MSM.953235
Med Sci Monit 2026; 32:e953235
Figure 2 Candidate variable screening diagram based on BorutaThe figure illustrates the results of feature selection using the Boruta algorithm. Feature variables are displayed along the x-axis, and their importance scores are shown on the y-axis. Green boxes represent features confirmed as important (“Confirmed”); red boxes denote features rejected as unimportant (“Rejected”); and blue boxes indicate the minimum, median, and maximum importance of random shadow features used for comparison (“Shadow”). Abbreviations: AID, autoimmune disease; APCD, adverse pregnancy and childbirth history; CVD, cardiovascular disease; GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy; IVF-ET, in vitro fertilization and embryo transfer; PM, primipara or multipara.