The project emerges as an opportunity to explore new construction techniques and materials, reinforcing its experimental profile and total creative freedom. With the façade facing the lush greenery of the avenue, the residence harmoniously integrates into the environment, reflecting a design that prioritizes both functionality and connection with nature.
© Rafa D'AndreaAnother highlight of the project is the natural ventilation system provided by the lantern
a tubular steel structure located between the technical floor and the upper floor
This ingenious solution utilizes convection to keep the house ventilated and comfortable while protecting the interior of the residence from rainwater ingress
You'll now receive updates based on what you follow
Personalize your stream and start following your favorite authors
If you have done all of this and still can't find the email
Metrics details
Low birth weight (LBW) is a critical factor linked to neonatal morbidity and mortality
Early prediction is essential for timely interventions
This study aimed to develop and evaluate predictive models for LBW using machine learning algorithms
We analyzed data from 1,579 pregnant women enrolled in the Araraquara Cohort
Predictor variables included maternal sociodemographic
were trained using an 80/20 train-test split and 10-fold cross-validation
the Synthetic Minority Over-sampling Technique (SMOTE) was applied
Model performance was assessed using metrics such as area under the receiver operating characteristic curve (AUROC)
Variable importance was evaluated using Shapley values
Maternal gestational age was the most influential predictor
followed by marital status and prenatal care frequency
Shapley analysis provided interpretable insights into variable contributions
supporting the clinical applicability of the models
proved to be an effective approach for predicting LBW
XGBoost stood out as the most accurate model
but Catboost and Random Forest also provided solid results
These models can be applied to identify high-risk pregnancies
improving perinatal outcomes through early interventions
The application of these algorithms can enhance the precision and reliability of predictions
contributing to early prevention and intervention strategies for fetal growth issues
The aim of this study is to develop a predictive model for LBW in pregnant women using machine learning algorithms
By incorporating advanced techniques such as SMOTE and Shapley values
we aim to improve the accuracy and interpretability of predictions
thereby enabling more effective strategies to address fetal growth problems early and improve neonatal outcomes
The sample included women with a gestational age of ≤ 19 weeks who received prenatal care at Basic Health Units in Araraquara
The pregnant women were followed quarterly throughout their pregnancy until the birth of their children between 2017 and 2022
Women with twin pregnancies and those who experienced miscarriage were excluded
only data from the pregnancy were considered
The data used in this study were accessed between May 2024 and July 2024
The outcome of low birth weight was analyzed based on the dichotomous classification of birth weight, defined as low birth weight: < 2500 g and normal weight: ≥ 2500 g. The predictor variables are illustrated in Table 1
The study was approved by the Research Ethics Committee with Human Subjects at the School of Public Health
under protocol number CAEE: 59787216.2.0000.5421
All participants signed the Informed Consent Form before participating
The participants were informed about the objectives of the study
ensuring that their participation was entirely voluntary
All participants provided informed consent
consistent with the principles outlined in the Helsinki Declaration
An example of the workflow diagram for classifying birth weight adequacy
The data analysis was conducted using Python (version 3.9)
with key libraries including pandas (1.3.5) and numpy (1.21.4) for data preprocessing
the official libraries for CatBoost (1.0.4)
and LightGBM (3.3.2) were used for these respective algorithms
According to Table 2
the maternal characteristics of 1,579 pregnant women from the Araraquara cohort were evaluated
The women had an average age of 28.4 years
a pre-pregnancy body mass index (BMI) of 24.7 kg/m²
Most women (88.4%) had an education level equal to or greater than 8 years
and 87.7% were married or in a stable relationship
with most having a family income of R$563 and being non-smokers
A total of 11.3% of the women had a urinary tract infection during pregnancy
CRP (C-reactive protein) levels were 3.3 mg/L (interquartile range: 1.4–7.8)
and HOMA (homeostasis model assessment) values were 2.9 units (interquartile range: 1.3–6.1)
shows the distribution of low birth weight in the Araraquara cohort
where 1,309 (91.2%) had normal birth weight
we evaluated the predictive capacity of various ML models
The analysis focused on multiple performance metrics such as AUROC
and the MCC to provide a comprehensive evaluation of the ability of these models to classify cases of low birth weight
Performance by ROC curve of ML (Random Forest
Predictors of LBW (Shapley Variable Importance) for the best model - XGBoost
Figure 3
shows the importance of the predictors for LBW using Shapley values for the best-performing model
The most important variable identified was gestational age
standing out as the factor with the most influence on predicting low birth weight
maternal marital status and the absence of regular physical activity during pregnancy were significant predictors
Other factors contributing substantially included maternal race
underscoring the importance of socioeconomic and behavioral variables
Variables such as smoking and alcohol consumption during pregnancy
appeared with less relevance compared to the previously mentioned factors
illustrates the strength of variable contributions to the prediction of LBW using Shapley values in the XGBoost model
provide a detailed view of how each factor contributes to increasing or reducing the risk of LBW
The use of ML in LBW prediction is gaining traction
as these models demonstrate significant potential across various clinical and public health applications
The use of Shapley values in our study provided additional insights into the interpretability of the models
allowing for a more granular understanding of how each variable contributes to the prediction of LBW
This aspect is particularly important in clinical applications where transparency and explainability of the models are crucial for their adoption by healthcare professionals
the Shapley analysis in our study indicated that gestational weight gain had the strongest influence on LBW predictions
followed by maternal race and prenatal care visits
These findings mirror global trends in LBW prediction
and prenatal care are recognized as key determinants of birth outcomes
Their study found that socioeconomic factors
including maternal education and access to prenatal care
significantly contributed to the risk of LBW
similar to our results from the Araraquara cohort
Both studies emphasize the importance of early identification of high-risk pregnancies through predictive models
especially in resource-limited settings where timely interventions can significantly improve neonatal outcomes
While the results of this study are promising
the cohort used in this study was from a specific region in Brazil
which may limit the generalizability of the findings to other populations
Future research should aim to validate these models across different regions and populations to ensure their broader applicability
while our models demonstrated high predictive accuracy
further research is needed to assess their integration into clinical workflows and their potential impact on perinatal care
future studies could explore the use of these models in combination with mobile health (mHealth) technologies to improve prenatal care in low-resource settings
although we employed robust techniques such as SMOTE to handle class imbalance and Shapley values to interpret model predictions
the models still require validation in real-world clinical environments
The practical deployment of ML models in healthcare settings involves challenges related to data privacy
which should be thoroughly addressed before these models can be widely adopted
This study successfully developed and evaluated machine learning models
for predicting low birth weight in neonates
The XGBoost model demonstrated the highest predictive performance
with excellent discrimination between neonates at risk of LBW and those with normal birth weight
The application of these models in clinical practice has the potential to improve early detection of high-risk pregnancies
enabling timely and personalized interventions that could significantly improve neonatal outcomes
Given the increasing global focus on maternal and neonatal health
these findings hold important implications for both clinical practice and public health policy
The integration of machine learning models into prenatal care systems could offer a transformative approach to preventing adverse birth outcomes
particularly in low-resource settings where LBW remains a critical challenge
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. The code developed for constructing the algorithms along, is available on Github (https://github.com/Audency/Predictors-of-Low-Birth-Weight-using-Machine-Laerning-.git.)
Risk factors for low birth weight in hospitals of North Wello zone
Evaluating association of maternal nutritional status with neonatal birth weight in term pregnancies: A Cross-Sectional study with unexpected outcomes
Revista Brasileira De Cineantropometria E Desempenho Humano
O Crescimento e desenvolvimento Frente à prematuridade e Baixo peso Ao nascer
Low birth weight and its associated factors
The limits of small-for-gestational-age as a high-risk category
Gätjens I, Fedde S, Schmidt SCE, Hasler M, Plachta-Danielzik S, Müller MJ, et al. Relationship between birth weight, early growth rate and body composition in 5 to 7-year-old children. Obes Facts. 2022. https://doi.org/10.1159/000522509
Perinatal health predictors using artificial intelligence: A review
A systematic review of the digital interventions for fighting COVID-19: the Bangladesh perspective
Machine learning in ‘big data’: handle with care
Towards deep phenotyping pregnancy: a systematic review on artificial intelligence and machine learning methods to improve pregnancy outcomes
Building a predictive model of low birth weight in low- and middle-income countries: a prospective cohort study
Prediction of low birth weight from fetal ultrasound and clinical characteristics: a comparative study between a low- and middle-income and a high-income country
SMOTE: synthetic minority Over-sampling technique
Machine learning-based approach for predicting low birth weight
Issue of data imbalance on low birthweight baby outcomes prediction and associated risk factors identification: establishment of benchmarking key machine learning models with data rebalancing strategies
The impact of gestational weight gain on fetal and neonatal outcomes: the Araraquara cohort study
Predictive modeling of gestational weight gain: a machine learning multiclass classification study
CatBoost: unbiased boosting with categorical features
In: Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining
Lightgbm: A highly efficient gradient boosting decision tree
Making a science of model search: Hyperparameter optimization in hundreds of dimensions for vision architectures
In: International conference on machine learning
Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): explanation and elaboration
A machine-learning–based algorithm improves prediction of preeclampsia-associated adverse outcomes
Machine learning for fetal growth prediction
Machine learning methods for preterm birth prediction: A review
Machine learning to predict pregnancy outcomes: a systematic review
synthesizing framework and future research agenda
Deep neural networks and tabular data: A survey
Tabular data: deep learning is not all you need
When Do Neural Nets Outperform Boosted Trees on Tabular Data
Predicting risks of low birth weight in Bangladesh with machine learning
Download references
The authors gratefully acknowledge the professionals
and graduate students who collaborated in the data collection for the Araraquara cohort
This study was supported by the São Paulo Research Foundation (FAPESP)(grant number 2015/03333–6)
Audêncio Victor has received a scholarship from São Paulo Research Foundation (FAPESP) (grant number 2023/07936-3)
Faculdade de Saúde Pública- USP Avenida Doutor Arnaldo
Administração e Contabilidade de Ribeirão Preto
SPX: visualization and writing - original draft
All authors contributed to the article and approved the submitted version
The research received ethical approval from the Research Ethics Committee with Human Subjects at the School of Public Health
before the commencement of data collection
as per protocol CAEE: 59787216.2.0000.5421 and opinion number 1.885.874
The authors declare no competing interests
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations
Below is the link to the electronic supplementary material
Download citation
DOI: https://doi.org/10.1186/s12884-025-07351-3
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
RIO DE JANEIRO (LNB) - Flamengo are in a rich vein of form heading into the Liga Sudamericana.\r\n\r\nThe competition tips off on February 17 and Flamengo, who lost out to Universo/BRB in their battle to reach the FIBA Americas League Final Four, have rebounded to win all four of their top flight games in Brazil.\r\n\r\nIn Sao Paulo, meanwhile, Lupo/Araraquara ...
HomeNewsBRA – Flamengo flying in Brazil but Araraquara upset Universo/BRB FIBA BasketballBRA – Flamengo flying in Brazil but Araraquara upset Universo/BRB RIO DE JANEIRO (LNB) - Flamengo are in a rich vein of form heading into the Liga Sudamericana
The competition tips off on February 17 and Flamengo
who lost out to Universo/BRB in their battle to reach the FIBA Americas League Final Four
have rebounded to win all four of their top flight games in Brazil
Connecting decision makers to a dynamic network of information
Bloomberg quickly and accurately delivers business and financial information
EMTs arrive with a patient at Vila Xavier Hospital in Araraquara
The city has seen more Covid-19 deaths in the first two months of the year than in all of 2020.
Three adjacent cities with three different approaches illustrate the despair
Parece que a página que você está procurando não está disponível.
Connecting decision makers to a dynamic network of information, people and ideas, Bloomberg quickly and accurately delivers business and financial information, news and insight around the world
2015 at 9:08 AM ESTBookmarkSaveLock This article is for subscribers only
The tang of Jose Luis Cutrale’s orange juice factory permeates the air in the Brazilian town of Araraquara
For visitors who make the 173-mile trek northwest from Sao Paulo
the aroma is a welcome respite from the stench that normally emanates from the city’s sewage-choked Tiete river
the Sucocitrico Cutrale Ltda plant reminds them of home
This website is using a security service to protect itself from online attacks
The action you just performed triggered the security solution
There are several actions that could trigger this block including submitting a certain word or phrase
You can email the site owner to let them know you were blocked
Please include what you were doing when this page came up and the Cloudflare Ray ID found at the bottom of this page.