Predicting fetal hypoxia using common spatial pattern and machine learning from cardiotocography signals

Alsaggaf, Wafaa; Cömert, Zafer; Polat, Kemal; Brdesee, Hani; Toğaçar, Mesut

doi:10.1016/j.apacoust.2020.107429

Cited by 34 publications

(17 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The models’ performance were evaluated in terms of the Area Under the receiver operating Curve (AUC) and, once the optimal threshold was identified by Youden’s index on ROC curve [ 23 ], standard metrics, such as accuracy, sensitivity and specificity were also computed: where TP and TN stand for True Positive (number of IDE cases correctly classified) and True Negative (number of non-IDE cases correctly classified), while FP (number of non-IDE cases identified as IDE cases) and FN (number of IDE cases identified as non-IDE cases are False Positive and False Negative ones, respectively. In the case of performance evaluation of the ensemble model, patients for which no prediction was given were not counted in any of the four categories (TP, TN, FP, FN), thus the number of “no answers” were computed [ 24 , 25 ].…”

Section: Methodsmentioning

confidence: 99%

“…where TP and TN stand for True Positive (number of IDE cases correctly classified) and True Negative (number of non-IDE cases correctly classified), while FP (number of non-IDE cases identified as IDE cases) and FN (number of IDE cases identified as non-IDE cases are False Positive and False Negative ones, respectively. In the case of performance evaluation of the ensemble model, patients for which no prediction was given were not counted in any of the four categories (TP, TN, FP, FN), thus the number of "no answers" were computed [24,25].…”

Section: Plos Onementioning

confidence: 99%

See 1 more Smart Citation

A machine learning ensemble approach for 5- and 10-year breast cancer invasive disease event classification

Massafra

Comes²,

Bove³

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

Designing targeted treatments for breast cancer patients after primary tumor removal is necessary to prevent the occurrence of invasive disease events (IDEs), such as recurrence, metastasis, contralateral and second tumors, over time. However, due to the molecular heterogeneity of this disease, predicting the outcome and efficacy of the adjuvant therapy is challenging. A novel ensemble machine learning classification approach was developed to address the task of producing prognostic predictions of the occurrence of breast cancer IDEs at both 5- and 10-years. The method is based on the concept of voting among multiple models to give a final prediction for each individual patient. Promising results were achieved on a cohort of 529 patients, whose data, related to primary breast cancer, were provided by Istituto Tumori “Giovanni Paolo II” in Bari, Italy. Our proposal greatly improves the performances returned by the baseline original model, i.e., without voting, finally reaching a median AUC value of 77.1% and 76.3% for the IDE prediction at 5-and 10-years, respectively. Finally, the proposed approach allows to promote more intelligible decisions and then a greater acceptability in clinical practice since it returns an explanation of the IDE prediction for each individual patient through the voting procedure.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Plos Onementioning

confidence: 99%

A machine learning ensemble approach for 5- and 10-year breast cancer invasive disease event classification

Massafra

Comes²,

Bove³

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…The FHR signal limitations ( H & L ) determined from a truncated FHR signal are indicated in Fig. 6 ( Alsaggaf et al, 2020 ; Krupa, 2010 ).…”

Section: Methodsmentioning

confidence: 99%

“…However, their problem is that they also require a reference signal, and the QRS complex may suffer from interference with the parameter estimation, where the QRS Complex One single heartbeat is equivalent to the depolarization of the right and left ventricles (lower cardiac chambers) (lower heart chambers), the (Q-wave = initial negative deflection), the (R-wave = initial positive deflection, the S-wave = second negative deflection). The researchers ( Alsaggaf et al, 2020 ) presented a method based on a triple filter as well as dual covers characteristic chosen approaches and machine learning models, in addition to models that were assessed according to a feature set of high dimension secured from a publicly available Czech Technical University source and database obtained from the University Hospital in Brno (UHB) intrapartum cardiotocography. Due to the various division criteria, this poses a considerably challenging issue.…”

Section: Literature Reviewmentioning

confidence: 99%

Intrapartum cardiotocography trace pattern pre-processing, features extraction and fetal health condition diagnoses based on RCOG guideline

Al-Yousif

Najm

Talab

et al. 2022

PeerJ Computer Science

View full text Add to dashboard Cite

Context The computerization of both fetal heart rate (FHR) and intelligent classification modeling of the cardiotocograph (CTG) is one of the approaches that are utilized in assisting obstetricians in conducting initial interpretation based on (CTG) analysis. CTG tracing interpretation is crucial for the monitoring of the fetal status during weeks into the pregnancy and childbirth. Most contemporary studies rely on computer-assisted fetal heart rate (FHR) feature extraction and CTG categorization to determine the best precise diagnosis for tracking fetal health during pregnancy. Furthermore, through the utilization of a computer-assisted fetal monitoring system, the FHR patterns can be precisely detected and categorized. Objective The goal of this project is to create a reliable feature extraction algorithm for the FHR as well as a systematic and viable classifier for the CTG through the utilization of the MATLAB platform, all the while adhering to the recognized Royal College of Obstetricians and Gynecologists (RCOG) recommendations. Method The compiled CTG data from spiky artifacts were cleaned by a specifically created application and compensated for missing data using the guidelines provided by RCOG and the MATLAB toolbox after the implemented data has been processed and the FHR fundamental features have been extracted, for example, the baseline, acceleration, deceleration, and baseline variability. This is followed by the classification phase based on the MATLAB environment. Next, using the guideline provided by the RCOG, the signals patterns of CTG were classified into three categories specifically as normal, abnormal (suspicious), or pathological. Furthermore, to ensure the effectiveness of the created computerized procedure and confirm the robustness of the method, the visual interpretation performed by five obstetricians is compared with the results utilizing the computerized version for the 150 CTG signals. Results The attained CTG signal categorization results revealed that there is variability, particularly a trivial dissimilarity of approximately (+/−4 and 6) beats per minute (b.p.m.). It was demonstrated that obstetricians’ observations coincide with algorithms based on deceleration type and number, except for acceleration values that differ by up to (+/−4). Discussion The results obtained based on CTG interpretation showed that the utilization of the computerized approach employed in infirmaries and home care services for pregnant women is indeed suitable. Conclusions The classification based on CTG that was used for the interpretation of the FHR attribute as discussed in this study is based on the RCOG guidelines. The system is evaluated and validated by experts based on their expert opinions and was compared with the CTG feature extraction and classification algorithms developed using MATLAB.

show abstract

“…Machine learning (ML) is a branch of artificial intelligence whose aim is to recognize hidden patterns automatically from data. Recently, it has been frequently used to deal with biomedical issues in several contexts: cardiology [7,8], fetal monitoring [9][10][11], medical imaging analysis [12,13], oncology [14][15][16] and in several other medical specialties [17][18][19]. Problems regarding regression or classification have been solved by applying state-of-art algorithms which proved to help clinicians in handling difficult tasks.…”

Section: Introductionmentioning

confidence: 99%

Bidimensional and Tridimensional Poincaré Maps in Cardiology: A Multiclass Machine Learning Study

et al. 2022

View full text Add to dashboard Cite

Heart rate is a nonstationary signal and its variation may contain indicators of current disease or warnings about impending cardiac diseases. Hence, heart rate variation analysis has become a noninvasive tool to further study the activities of the autonomic nervous system. In this scenario, the Poincaré plot analysis has proven to be a valuable tool to support cardiac diseases diagnosis. The study’s aim is a preliminary exploration of the feasibility of machine learning to classify subjects belonging to five cardiac states (healthy, hypertension, myocardial infarction, congestive heart failure and heart transplanted) using ten unconventional quantitative parameters extracted from bidimensional and three-dimensional Poincaré maps. Knime Analytic Platform was used to implement several machine learning algorithms: Gradient Boosting, Adaptive Boosting, k-Nearest Neighbor and Naïve Bayes. Accuracy, sensitivity and specificity were computed to assess the performances of the predictive models using the leave-one-out cross-validation. The Synthetic Minority Oversampling technique was previously performed for data augmentation considering the small size of the dataset and the number of features. A feature importance, ranked on the basis of the Information Gain values, was computed. Preliminarily, a univariate statistical analysis was performed through one-way Kruskal Wallis plus post-hoc for all the features. Machine learning analysis achieved interesting results in terms of evaluation metrics, such as demonstrated by Adaptive Boosting and k-Nearest Neighbor (accuracies greater than 90%). Gradient Boosting and k-Nearest Neighbor reached even 100% score in sensitivity and specificity, respectively. The most important features according to information gain are in line with the results obtained from the statistical analysis confirming their predictive power. The study shows the proposed combination of unconventional features extracted from Poincaré maps and well-known machine learning algorithms represents a valuable approach to automatically classify patients with different cardiac diseases. Future investigations on enriched datasets will further confirm the potential application of this methodology in diagnostic.

show abstract

Predicting fetal hypoxia using common spatial pattern and machine learning from cardiotocography signals

Cited by 34 publications

References 66 publications

A machine learning ensemble approach for 5- and 10-year breast cancer invasive disease event classification

A machine learning ensemble approach for 5- and 10-year breast cancer invasive disease event classification

Intrapartum cardiotocography trace pattern pre-processing, features extraction and fetal health condition diagnoses based on RCOG guideline

Bidimensional and Tridimensional Poincaré Maps in Cardiology: A Multiclass Machine Learning Study

Contact Info

Product

Resources

About