Hypertensive Disorders of Pregnancy: Kurtosis-Based Classification of Fetal Doppler Ultrasound Signals
| Autor: | Rayan Chaaban, Ayache Bouakaz, Amira Zaylaa, Walaa Issa |
|---|---|
| Přispěvatelé: | Department of Biomedical Engineering [Beirut, Lebanon] (LIU), Lebanese International University (LIU), Imagerie et cerveau (iBrain - Inserm U1253 - UNIV Tours ), Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Tours (UT), Neuroscience Research Center [Beirut, Lebanon] (Faculty of Medical Sciences), Lebanese University [Beirut] (LU), Faculty of Public Health-V [Beirut, Lebanon], Zaylaa, Amira, Université de Tours (UT)-Institut National de la Santé et de la Recherche Médicale (INSERM), Université de Tours-Institut National de la Santé et de la Recherche Médicale (INSERM) |
| Rok vydání: | 2019 |
| Předmět: |
[SDV.BIO]Life Sciences [q-bio]/Biotechnology
Intrauterine growth restriction K-means clustering 01 natural sciences 010309 optics 03 medical and health sciences [SDV.MHEP.PED] Life Sciences [q-bio]/Human health and pathology/Pediatrics 0302 clinical medicine 0103 physical sciences Feature (machine learning) Medicine Intrauterine Growth Restriction [SDV.IB] Life Sciences [q-bio]/Bioengineering [SDV.MHEP.PED]Life Sciences [q-bio]/Human health and pathology/Pediatrics Pregnancy Kurtosis [SDV.BIBS] Life Sciences [q-bio]/Quantitative Methods [q-bio.QM] business.industry Support Vector Machine Algorithm Ultrasound Fetal doppler Pattern recognition Hypertensive Disorders of Pregnancy medicine.disease [SDV.BIBS]Life Sciences [q-bio]/Quantitative Methods [q-bio.QM] [SDV.BIO] Life Sciences [q-bio]/Biotechnology 3. Good health Support vector machine Statistical classification Entropy Features Statistical Analysis [SDV.SPEE] Life Sciences [q-bio]/Santé publique et épidémiologie [SDV.IB]Life Sciences [q-bio]/Bioengineering [SDV.SPEE]Life Sciences [q-bio]/Santé publique et épidémiologie Artificial intelligence business 030217 neurology & neurosurgery |
| Zdroj: | The 5th International Conference on Advances in Biomedical Engineering, IEEE-EMBS The 5th International Conference on Advances in Biomedical Engineering, IEEE-EMBS, Oct 2019, Tripoli, Lebanon |
| DOI: | 10.1109/icabme47164.2019.8940240 |
| Popis: | International audience; Hypertensive Disorders of Pregnancy (HDP), a group of medical conditions occurring during pregnancy, have wide-reaching implications on the normal progress of 17% of pregnancies, leading to maternal and perinatal morbidity and mortality. One of the HDP complications is the Intrauterine Growth Restriction (IUGR). IUGR changes the behavior of any feature extracted from Fetal Heart Rates (FHRs). These features if well-selected improve the classification of IUGR. However, the choice of features was reliant on whether it's linear or nonlinear. Also, the classification algorithms such as, K-Means and Support Vector Machine (SVM) used to predict and classify biomedical signals were not optimal, and the best classification algorithm was not yet set. Our aim is to propose a new kurtosis-based combinations of features and explore their effect on HDP and IUGR classification from Doppler Ultrasound FHRs. Features extracted from FHRs were fed into K-means and SVM classification algorithms. The database comprised 50 normal and 50 IUGR FHRs. Results showed that the best extracted features were those based on kurtosis, and the best classification method was the SVM. The best combination result was 67% sensitive, 100% specific and 100% precise to the classification and detection of IUGR and thus HDP. A further future study could test additional combination of features and other classification-based methods to predict IUGR and thus HDP. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|