Dataset Information

Prediction of newborn's body mass index using nationwide multicenter ultrasound data: a machine-learning study.

ABSTRACT:

Background

This study introduced machine learning approaches to predict newborn's body mass index (BMI) based on ultrasound measures and maternal/delivery information.

Methods

Data came from 3159 obstetric patients and their newborns enrolled in a multi-center retrospective study. Variable importance, the effect of a variable on model performance, was used for identifying major predictors of newborn's BMI among ultrasound measures and maternal/delivery information. The ultrasound measures included biparietal diameter (BPD), abdominal circumference (AC) and estimated fetal weight (EFW) taken three times during the week 21 - week 35 of gestational age and once in the week 36 or later.

Results

Based on variable importance from the random forest, major predictors of newborn's BMI were the first AC and EFW in the week 36 or later, gestational age at delivery, the first AC during the week 21 - the week 35, maternal BMI at delivery, maternal weight at delivery and the first BPD in the week 36 or later. For predicting newborn's BMI, linear regression (2.0744) and the random forest (2.1610) were better than artificial neural networks with one, two and three hidden layers (150.7100, 154.7198 and 152.5843, respectively) in the mean squared error.

Conclusions

This is the first machine-learning study with 64 clinical and sonographic markers for the prediction of newborns' BMI. The week 36 or later is the most effective period for taking the ultrasound measures and AC and EFW are the best predictors of newborn's BMI alongside gestational age at delivery and maternal BMI at delivery.

SUBMITTER: Lee KS

PROVIDER: S-EPMC7927215 | biostudies-literature | 2021 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Prediction of newborn's body mass index using nationwide multicenter ultrasound data: a machine-learning study.

Lee Kwang-Sig KS Kim Ho Yeon HY Lee Se Jin SJ Kwon Sung Ok SO Na Sunghun S Hwang Han Sung HS Park Mi Hye MH Ahn Ki Hoon KH

BMC pregnancy and childbirth 20210302 1

<h4>Background</h4>This study introduced machine learning approaches to predict newborn's body mass index (BMI) based on ultrasound measures and maternal/delivery information.<h4>Methods</h4>Data came from 3159 obstetric patients and their newborns enrolled in a multi-center retrospective study. Variable importance, the effect of a variable on model performance, was used for identifying major predictors of newborn's BMI among ultrasound measures and maternal/delivery information. The ultrasound ...[more]

PMID: 33653299

Similar Datasets

Project description:There has been no machine learning study with a rich collection of clinical, sonographic markers to compare the performance measures for a variety of newborns' weight-for-height indicators. This study compared the performance measures for a variety of newborns' weight-for-height indicators based on machine learning, ultrasonographic data and maternal/delivery information. The source of data for this study was a multi-center retrospective study with 2949 mother-newborn pairs. The mean-squared-error-over-variance measures of five machine learning approaches were compared for newborn's weight, newborn's weight/height, newborn's weight/height2 and newborn's weight/hieght3. Random forest variable importance, the influence of a variable over average node impurity, was used to identify major predictors of these newborns' weight-for-height indicators among ultrasonographic data and maternal/delivery information. Regarding ultrasonographic fetal biometry, newborn's weight, newborn's weight/height and newborn's weight/height2 were better indicators with smaller mean-squared-error-over-variance measures than newborn's weight/height3. Based on random forest variable importance, the top six predictors of newborn's weight were the same as those of newborn's weight/height and those of newborn's weight/height2: gestational age at delivery time, the first estimated fetal weight and abdominal circumference in week 36 or later, maternal weight and body mass index at delivery time, and the first biparietal diameter in week 36 or later. These six predictors also ranked within the top seven for large-for-gestational-age and the top eight for small-for-gestational-age. In conclusion, newborn's weight, newborn's weight/height and newborn's weight/height2 are more suitable for ultrasonographic fetal biometry with smaller mean-squared-error-over-variance measures than newborn's weight/height3. Machine learning with ultrasonographic data would be an effective noninvasive approach for predicting newborn's weight, weight/height and weight/height2.

Project description:ImportanceBody mass index (BMI) is used to diagnose obesity in adolescents worldwide, despite evidence that weight does not scale with height squared in adolescents. To account for this, health care providers diagnose obesity using BMI percentiles for each age (BMI z scores), but this does not ensure that BMI is accurate in adolescents.ObjectiveTo compare the accuracy of BMI vs other body fat indices of the form body mass divided by heightn in estimating body fat levels in adolescents.Design, setting, and participantsCross-sectional data from the 1999 to 2006 US National Health and Nutrition Examination Survey were analyzed between September 2015 and December 2016.Main outcomes and measuresDual-energy x-ray absorptiometry and anthropometric data were used to determine changes in body fat levels, body proportions, and the scaling relationships among body mass, height, and percent body fat. To assess the merits of each adiposity index, 3 criteria were used: stability with age, accuracy in estimating percent body fat, and accuracy in classifying adolescents as overweight vs normal weight.ResultsParticipants included 2285 non-Hispanic white participants aged 8 to 29 years. Percent body fat varied with both age and height during adolescence, invalidating the standard weight-to-height regression as the way of finding the optimal body fat index. Because the correct regression model (percent body fat is proportional to mass divided by heightn) suggested that percent body fat scales to height with an exponent closer to 3, we therefore focused on the tri-ponderal mass index (TMI; mass divided by height cubed) as an alternative to BMI z scores. For ages 8 to 17 years, TMI yielded greater stability with age and estimated percent body fat better than BMI (R2?=?0.64 vs 0.38 in boys and R2?=?0.72 vs 0.66 in girls). Moreover, TMI misclassified adolescents as overweight vs normal weight less often than BMI z scores (TMI, 8.4%; 95% CI, 7.3%-9.5% vs BMI, 19.4%; 95% CI, 17.8%-20.0%; P?<?.001) and performed equally as well as updated BMI percentiles derived from the same data set (TMI, 8.4%; 95% CI, 7.3%-9.5% vs BMI, 8.0%; 95% CI, 6.9%-9.1%; P?=?.62).Conclusions and relevanceThe tri-ponderal mass index estimates body fat levels more accurately than BMI in non-Hispanic white adolescents aged 8 to 17 years. Moreover, TMI diagnoses adolescents as overweight more accurately than BMI z scores and equally as well as updated BMI percentiles but is much simpler to use than either because it does not involve complicated percentiles. Taken together, it is worth considering replacing BMI z scores with TMI to estimate body fat levels in adolescents.

Dataset Information

Prediction of newborn's body mass index using nationwide multicenter ultrasound data: a machine-learning study.

Background

Methods

Results

Conclusions

Publications

Prediction of newborn's body mass index using nationwide multicenter ultrasound data: a machine-learning study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets