Artificial intelligence in human reproduction

Archives of Medical Research, Journal Year: 2024, Volume and Issue: 55(8), P. 103131 - 103131

Published: Nov. 29, 2024

Language: Английский

---

Distinguishing preeclampsia using the falling scaled slope (FSS) --- a novel photoplethysmographic morphological parameter

Hypertension in Pregnancy, Journal Year: 2023, Volume and Issue: 42(1)

Published: June 19, 2023

Background Preeclampsia (PE) presence could lead to hemodynamic changes. Previous research suggested that morphological parameters based on photoplethysmographic pulse waves (PPGW) help diagnose PE.Aim To investigate the performance of a novel PPGPW-based parameter, falling scaled slope (FSS), in distinguishing PE. advantages machine learning algorithm over conventional statistical methods analysis.Methods Eighty-one pieces PPGPW data were acquired for study (PE, n = 44; normotensive, 37). The FSS values calculated and used construct PE classifier using K-nearest neighbors (KNN) algorithm. A predicted state varying from 0 1 was also calculated. classifier's evaluated ROC AUC. comparison conducted with previously published models.Result Compared previous parameters, showed better an AUC value 0.924, best threshold 0.498 predict sensitivity 84.1% specificity 89.2%. As analysis method, training KNN had advantage 0.878 0.749, respectively.Conclusion result indicated might be effective tool identifying Moreover, further improve performance.

Language: Английский

---

Machine Learning Techniques for Predicting Pregnancy Complications

Advances in computational intelligence and robotics book series, Journal Year: 2023, Volume and Issue: unknown, P. 116 - 125

Published: Sept. 25, 2023

Machine learning is employed extensively in healthcare, prediction, diagnosis, and as a technique of establishing priority. Artificial intelligence widely used the medical industry. There are variety tools disciplines obstetrics childcare that use machine techniques. The goal current chapter to examine research development views employ approaches identify different complications during delivery. common such gestational diabetes mellitus, preeclampsia, stillbirth, depression anxiety, preterm labor, high blood pressure, miscarriage were explored this chapter. It investigated synthesized picture features utilized, types features, data sources, its characteristics; it analyzed adopted algorithms their performances; gave summary employed. Eventually, results review helped create conceptual framework for improving maternal healthcare system based on learning.

Language: Английский

---

Deep Survival Analysis for Interpretable Time-Varying Prediction of Preeclampsia Risk

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 20, 2024

Abstract Objective Survival analysis is widely utilized in healthcare to predict the timing of disease onset. Traditional methods survival are usually based on Cox Proportional Hazards model and assume proportional risk for all subjects. However, this assumption rarely true most diseases, as underlying factors have complex, non-linear, time-varying relationships. This concern especially relevant pregnancy, where pregnancy-related complications, such preeclampsia, varies across gestation. Recently, deep learning models shown promise addressing limitations classical models, novel allow non-proportional handling, capturing nonlinear relationships, navigating complex temporal dynamics. Methods We present a methodology preeclampsia during pregnancy investigate associated clinical factors. retrospective dataset including 66,425 pregnant individuals who delivered two tertiary care centers from 2015-2023. modeled by modifying DeepHit, model, which leverages neural network architecture capture relationships between covariates pregnancy. applied time series k-means clustering DeepHit’s normalized output investigated interpretability using Shapley values. Results demonstrate that DeepHit can effectively handle high-dimensional data evolving hazards over with performance similar achieving an area under curve (AUC) 0.78 both models. The outperformed traditional identifying time-varied trajectories providing insights early individualized intervention. K-means resulted patients delineating into low-risk, early-onset, late-onset groups— notably, each those has distinct Conclusion work demonstrates application prediction risk. Our results highlight advantage compared personalized trajectory demonstrating potential generate interpretable meaningful applications medicine.

Language: Английский

---

Explainable artificial hydrocarbon networks classifier applied to preeclampsia

Information Sciences, Journal Year: 2024, Volume and Issue: 670, P. 120556 - 120556

Published: April 8, 2024

Explainability is crucial in domains where system decisions have significant implications for human trust black-box models. Lack of understanding regarding how these are made hinders the adoption so-called clinical decision support systems. While neural networks and deep learning methods exhibit impressive performance, they remain less explainable than white-box approaches. Artificial Hydrocarbon Networks (AHN) an effective model that can be used to critical if accompanied by explainability mechanisms instill confidence among clinicians. In this paper, we present a use case involving global local explanations AHN models, provided with automatic procedure eXplainable (XAHN). We apply XAHN preeclampsia prognosis, enabling interpretability within accurate model. Our approach involves training suitable using cross-validation ten repetitions, followed comparative analysis against four well-known machine techniques. Notably, outperformed others, achieving F1-score 74.91%. Additionally, assess efficacy our explainer through survey applied clinicians, evaluating goodness satisfaction explanations. To best knowledge, work represents one earliest attempts address challenge prediction.

Language: Английский

---

A Prospective Study on Risk Prediction of Preeclampsia Using Bi-Platform Calibration and Machine Learning

International Journal of Molecular Sciences, Journal Year: 2024, Volume and Issue: 25(19), P. 10684 - 10684

Published: Oct. 4, 2024

Preeclampsia is a pregnancy syndrome characterized by complex symptoms which cause maternal and fetal problems deaths. The aim of this study to achieve preeclampsia risk prediction early in Xinjiang, China, based on the placental growth factor measured using SiMoA or Elecsys platform. A novel reliable calibration modeling method missing data imputing are proposed, different strategies used adapt small samples, training data, test independent features, dependent feature pairs. Multiple machine learning algorithms were applied train models various datasets, such as single-platform versus bi-platform plus non-early real augmented data. It was found that combination two types mono-platform could improve performance, enhance performance when limited available. Additionally, inclusion resulted achieving high but unstable performance. significantly reduced incidence region from 7.2% 2.0%, mortality rate 0%.

Language: Английский

---

ACIDA-IA: Addressing Class Imbalance Does not Always Improve Accuracy: a Study of Bayesian Networks for Predicting Preeclampsia

Published: July 27, 2024

Language: Английский

---

Early Prediction of Hypertensive Disorders of Pregnancy Using Machine Learning and Medical Records from the First and Second Trimesters

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 25, 2024

Hypertensive disorders of pregnancy (HDPs) remain a major challenge in maternal health. Early prediction HDPs is crucial for timely intervention. Most existing predictive machine learning (ML) models rely on costly methods like blood, urine, genetic tests, and ultrasound, often extracting features from data gathered throughout pregnancy, delaying This study developed an ML model to identify HDP risk before clinical onset using affordable methods. Features were extracted blood pressure (BP) measurements, body mass index values (BMI) recorded during the first second trimesters, demographic information. We employed random forest classification its robustness ability handle complex datasets. Our dataset, large academic medical centers Atlanta, Georgia, United States (2010-2022), comprised 1,190 patients with 1,216 records collected trimesters. Despite limited number features, model's performance demonstrated strong accurately predict HDPs. The achieved F1- score, accuracy, positive value, area under receiver-operating characteristic curve 0.76, 0.72, 0.75, 0.78, respectively. In conclusion, was shown be effective capturing relevant patterns feature set necessary predicting Moreover, it can implemented simple devices, such as BP monitors weight scales, providing practical solution early low-resource settings proper testing validation. By improving detection HDPs, this approach potentially help management adverse outcomes.

Language: Английский

---

Preeclampsia prediction via machine learning: a systematic literature review

Health Systems, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 15

Published: Dec. 9, 2024

Preeclampsia, a life-threatening condition in late pregnancy, has unclear causes and risk factors. Machine learning (ML) offers promising approach for early prediction. This systematic review analyzes state-of-the-art studies on preeclampsia prediction using ML approaches. We reviewed articles published between January 1 2013 December 31 2023, from Google Scholar PubMed. Of 183 identified studies, 35 were selected based inclusion exclusion criteria. Our findings reveal that key predictive features commonly used machine models include age, number of pregnancies, body mass index, diabetes, hypertension, blood pressure. In contrast, factors such as medications, genetic data, clinical imaging considered less frequently. Random Forest, Support Vector Machine, Logistic Regression, Decision Tree, Naïve Bayes the most algorithms. Most conducted China USA, indicating geographic concentration. The field seen notable rise research, especially past two years, though many rely small datasets single hospitals. highlights need more diverse comprehensive research to enhance detection management preeclampsia.

Language: Английский

---

Early prediction of preeclampsia risk using artificial intelligence

AIP conference proceedings, Journal Year: 2024, Volume and Issue: 3188, P. 100060 - 100060

Published: Jan. 1, 2024

Language: Английский

---