Participants and variablesData consisted of 10,423 VLBW infants from the Korean Neonatal Network (KNN) database throughout January 2013-December 2017. The KNN began on April 2013 as a nationwide potential cohort registry of VLBW infants admitted or transferred to neonatal intensive care models throughout South Korea (It covers 74 neonatal intensive care models now). It collects the perinatal and neonatal information of VLBW infants primarily based on a standardized working procedure37.Five adverse birth outcomes had been thought of as binary dependent variables (no, sure), i.e., gestational age lower than 28 weeks (GA < 28), GA lower than 26 weeks (GA < 26), birth weight lower than 1000 g (BW < 1000), BW lower than 750 g (BW < 750) and SGA. Thirty-three predictors were included: sex—male (no, yes), birth-year (2013, 2014, 2015, 2016, 2017), birth-month (1, 2, …, 12), birth-season-spring (no, yes), birth-season-summer (no, yes), birth-season-autumn (no, yes), birth-season-winter (no, yes), number of fetuses (1, 2, 3, 4 or more), in vitro fertilization (no, yes), gestational diabetes mellitus (no, yes), overt diabetes mellitus (no, yes), pregnancy-induced hypertension (no, yes), chronic hypertension (no, yes), chorioamnionitis (no, yes), prelabor rupture of membranes (no, yes), prelabor rupture of membranes > 18 h (no, sure), antenatal steroid (no, sure), cesarean part (no, sure), oligohydramnios (no, sure), polyhydramnios (no, sure), maternal age (years), primipara (no, sure), maternal training (elementary, junior excessive, senior excessive, school or increased), maternal citizenship (Korea, Vietnam, China, Philippines, Japan, Cambodia, United States, Thailand, Mongolia, Other), paternal training (elementary, junior excessive, senior excessive, school or increased), paternal citizenship (Korea, Vietnam, China, Philippines, Japan, Cambodia, United States, Thailand, Mongolia, Other), single (no, sure), congenital an infection (no, sure), PM10 yr (PM10 for every year), PM10 month (PM10 for every birth-month), temperature common (for every year), temperature min (for every year) and temperature max (for every year). PM10 and temperature information got here from the Korea Meteorological Administration (PM10 https://data.kma.go.kr/data/climate/selectDustRltmList.do?pgmNo=68; temperature https://web.kma.go.kr/weather/climate/past_cal.jsp). The definition of every variable is given in Text S1, supplementary textual content.Statistical analysisThe synthetic neural community, the choice tree, the logistic regression, the Naïve Bayes, the random forest and the assist vector machine had been used for predicting preterm birth38,39,40,41,42,43. A call tree contains three components, i.e., a take a look at on an unbiased variable (intermediate be aware), an final result of the take a look at (department) and a worth of the dependent variable (terminal node). A naïve Bayesian classifier performs classification on the idea of Bayes’ theorem. Here, the theory states that the likelihood of the dependent variable given sure values of unbiased variables could be calculated primarily based on the chances of the unbiased variables given a sure worth of the dependent variable. A random forest is a set of many resolution timber, which make majority votes on the dependent variable (“bootstrap aggregation”). Let us take a random forest with 1000 resolution timber for example. Let us assume that unique information contains 10,000 members. Then, the coaching and take a look at of this random forest takes two steps. Firstly, new information with 10,000 members is created primarily based on random sampling with substitute, and a call tree is created primarily based on this new information. Here, some members in the unique information can be excluded from the brand new information and these leftovers are known as out-of-bag information. This course of is repeated 1000 occasions, i.e., 1000 new information are created, 1000 resolution timber are created and 1000 out-of-bag information are created. Secondly, the 1000 resolution timber make predictions on the dependent variable of each participant in the out-of-bag information, their majority vote is taken as their last prediction on this participant, and the out-of-bag error is calculated because the proportion of fallacious votes on all members in the out-of-bag data38,39.A assist vector machine estimates a bunch of “assist vectors”, that’s, a line or house known as “hyperplane”. The hyperplane separates information with the best hole between varied sub-groups. An synthetic neural community consists of “neurons”, data models mixed by way of weights. In common, the unreal neural community contains one enter layer, one, two or three intermediate layers and one output layer. Neurons in a earlier layer hyperlink with “weights” in the subsequent layer (Here, these weights denote the strengths of linkages between neurons in a earlier layer and their next-layer counterparts). This “feedforward” operation begins from the enter layer, runs by way of intermediate layers and ends in the output layer. Then, this course of is adopted by studying: These weights are up to date in line with their contributions for a spot between the precise and predicted last outputs. This “backpropagation” operation begins from the output layer, runs by way of intermediate layers and ends in the enter layer. The two processes are repeated till the efficiency measure reaches a sure limit38,39. Data on 10,423 observations with full data had been divided into coaching and validation units with a 70:30 ratio (7296 vs. 3127). Accuracy, a ratio of right predictions amongst 3127 observations, was employed as a regular for validating the fashions. Random forest variable significance, the contribution of a sure variable for the efficiency (GINI) of the random forest, was used for analyzing main predictors of adverse birth outcomes in VLBW infants together with PM10. The random cut up and analysis had been repeated 50 occasions then its common was taken for exterior validation44,45. R-Studio 1.3.959 (R-Studio Inc.: Boston, United States) was employed for the analysis throughout August 1, 2021–September 30, 2021.Ethic assertionThe KNN registry was accepted by the institutional evaluation board (IRB) at every collaborating hospital (IRB No. of Korea University Anam Hospital: 2013AN0115). Informed consent was obtained from the father or mother(s) of every toddler registered in the KNN. All strategies had been carried out in accordance with the IRB-approved protocol and in compliance with related pointers and rules.The names of the institutional evaluation board of the KNN collaborating hospitals had been as follows: The institutional evaluation board of Gachon University Gil Medical Center, The Catholic University of Korea Bucheon ST. Mary’s Hospital, The Catholic University of Korea Seoul ST. Mary’s Hospital, The Catholic University of Korea ST. Vincent’s Hospital, The Catholic University of Korea Yeouido ST. Mary’s Hospital, The Catholic University of Korea Uijeongbu ST. Mary’s Hospital, Gangnam Severance Hospital, Kyung Hee University Hospital at Gangdong, GangNeung Asan Hospital, Kangbuk Samsung Hospital, Kangwon National University Hospital, Konkuk University Medical Center, Konyang University Hospital, Kyungpook National University Hospital, Gyeongsang National University Hospital, Kyung Hee University Medical heart, Keimyung University Dongsan Medical Center, Korea University Guro Hospital, Korea University Ansan Hospital, Korea University Anam Hospital, Kosin University Gospel Hospital, National Health Insurance Service Iilsan Hospital, Daegu Catholic University Medical Center, Dongguk University Ilsan Hospital, Dong-A University Hospital, Seoul Metropolitan Government-Seoul National University Boramae Medical Center, Pusan National University Hospital, Busan ST. Mary’s Hospital, Seoul National University Bundang Hospital, Samsung Medical Center, Samsung Changwon Medical Center, Seoul National University Hospital, Asan Medical Center, Sungae Hospital, Severance Hospital, Soonchunhyang University Hospital Bucheon, Soonchunhyang University Hospital Seoul, Soonchunhyang University Hospital Cheonan, Ajou University Hospital, Pusan National University Children’s Hospital, Yeungnam University Hospital, Ulsan University Hospital, Wonkwang University School of Medicine & Hospital, Wonju Severance Christian Hospital, Eulji University Hospital, Eulji General Hospital, Ewha Womans University Medical.Center, Inje University Busan Paik Hospital, Inje University Sanggye Paik Hospital, Inje University Ilsan Paik Hospital, Inje University Haeundae Paik Hospital, Inha University Hospital, Chonnam National University Hospital, Chonbuk National University Hospital, Cheil General Hospital & Women’s Healthcare Center, Jeju National University Hospital, Chosun University Hospital, Chung-Ang University Hospital, CHA Gangnam Medical Center, CHA University, CHA Bundang Medical Center, CHA University, Chungnam National University Hospital, Chungbuk National University, Kyungpook National University Chilgok Hospital, Kangnam Sacred Heart Hospital, Kangdong Sacred Heart Hospital, Hanyang University Guri Hospital, and Hanyang University Medical Center.
https://www.nature.com/articles/s41598-022-16234-y
Machine learning-based risk factor analysis of adverse birth outcomes in very low birth weight infants
