A Novel Hybrid Model for Automated Analysis Of Cardiotocograms Using Machine Learning Algorithms

In this study, a new hybrid model was presented for the prediction of fetal state from fetal heart rate (FHR) and the uterine contraction (UC) signals obtained from cardiotocogram (CTG) recordings. CTG monitoring of FHR and uterine contractions during pregnancy and delivery provides information on the physiological status of the fetus to identify hypoxia. The precise information obtained from these records provides some ideas for interpreting the pathological condition of the fetus. Thus, with early intervention, it allows to prevent any negative situation that will occur in the fetus in the future. In this study, due to the importance of this subject, a new hybrid model was developed which can perform high rate accurate diagnosis using Machine Learning (ML) algorithms. In the hybrid model, 4 different ML algorithms (k Nearest Neighbors (k-NN), Decision Tree (DT), Naive Bayes (NB) and Support Vector Machine (SVM)) were used. While the diagnosis without the hybrid model was low, the improved hybrid model increased the accuracy by 34%. As a result of this hybrid model, 100% success was achieved for classification, test success, Accuracy, Sensitivity and Specificity with NB and DT ML algorithms.

PDF

___

[1] Tekin gündüz, S., kurtuldu, A., & Türkkan, I. Ş. I. K. (2017). Sağlık Hizmetlerinde Eşitsizlik ve Etik. Aksaray Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 8(4), 32-43.
[2] Zach, L., Chudáček, V., Kužílek, J., Spilka, J., Huptych, M., Burša, M., & Lhotská, L. (2011). Mobile CTG—Fetal heart rate assessment using Android platform. In Computing in Cardiology (pp. 249-252). IEEE.
[3] Andersson, S. U. S. A. N. N. E. (2011). Acceleration and deceleration detection and baseline estimation. Göteborg: Chalmers University of Technology.
[4] UnbornHeart, http://www.unbornheart.com/ , Accessed Date:14.5.2018.
[5] A. Pinas and E. Chandraharan, (2016). Continuous cardiotocography during labour: Analysis, classification and management, Best Practice & Research Clinical Obstetrics & Gynaecology, vol. 30, pp. 33-47.
[6] D. Ayres-de-Campos, C. Y. Spong, E. (2015). Chandraharan, and F. I. F. M. E. C. Panel, FIGO consensus guidelines on intrapartum fetal monitoring: Cardiotocography, Int J Gynaecol Obstet, vol. 131, pp. 13-24.
[7] J. Spilka, V. Chudáček, M. Koucký, L. Lhotská, M. Huptych, P. Janků, et al., (2012). Using nonlinear features for fetal heart rate classification, Biomedical Signal Processing and Control, vol. 7, pp. 350-357.
[8] R. Czabanski, M. Jezewski, K. Horoba, J. Jezewski, and J. Leski, (2016). Fuzzy Analysis of Delivery Outcome Attributes for Improving the Automated Fetal State Assessment, Applied Artificial Intelligence, vol. 30, pp. 556-571.
[9] H. Sahin and A. Subasi, (2015). Classification of the cardiotocogram data for anticipation of fetal risks using machine learning techniques, Applied Soft Computing, vol. 33, pp. 231-238.
[10] M. Romano, P. Bifulco, M. Ruffo, G. Improta, F. Clemente, M. Cesarelli, (2016). Software for computerised analysis of cardiotocographic traces, Comput. Methods Programs Biomed., 124, pp. 121-137.
[11] J. Kessler, D. Moster, S. Albrechfsen, (2014). Delay in intervention increases neonatal morbidity in births monitored with Cardiotocography and ST-waveform analysis, Acta Obs. Gynecol. Scand., 93 (2), pp. 175-181.
[12] P.A. Warrick, E.F. Hamilton, D. Precup, R.E. Kearney, (2010). Classification of normal and hypoxic fetuses from systems modeling of intrapartum cardiotocography, IEEE Trans. Biomed. Eng., 57 (4), pp. 771-779.
[13] A. Pinas, E. Chadraharan, (2016). Continuous Cardiotocography during labour: analysis, classification and management, Best. Pract. Res. Clin. Obstet. Gynaecol., 30, pp. 33-47.
[14] P.A. Warrick, E.F. Hamilton, D. Precup, R.E. Kearney (2010). Classification of normal and hypoxic fetuses from systems modeling of intrapartum cardiotocography, IEEE Trans. Biomed. Eng., 57 (4), pp. 771-779.
[15] H. Sahin, A. Subasi, (2015). Classification of the cardiotocogram data for anticipation of fetal risks using machine learning techniques, Appl. Soft Comput., 33, pp. 231-238
[16] H. Ocak, H.M. Ertunc, (2013). Prediction of fetal state from the cardiotocogram recordings using adaptive neuro-fuzzy inference systems, Neural comput. Appliations, 22 (6), pp. 1583-1589.
[17] C. Rotariu, A. Pasarica, G. Andruseac, H. Costin, D. (2014). Nemescu Automatic analysis of the fetal heart rate variability and uterine contractions, IEEE Electrical and Power Engineering, pp.[1] Tekin gündüz, S., kurtuldu, A., & Türkkan, I. Ş. I. K. (2017). Sağlık
[18] C. Rotariu, A. Pasarica, H. Costin, D,. Nemescu (2014). Spectral analysis of fetal heart rate variability associated with fetal acidosis and base deficit values, International Conference on Development and Application Systems, pp. 210-213.
[19] K. Maeda, (2014). Modalities of fetal evaluation to detect fetal compromise prior to the development of significant neurological damage, J. Obstet. adn Gynaecol. Res., 40 (10), pp. 2089-2094.
[20] H. Ocak, (2013). A medical decision support system based on support vector machines and the genetic algorithm for the evaluation of fetal well-being. J. Med. Syst., 37 (2), p. 9913.
[21] T. Peterek, P. Gajdos, P. Dohnalek, J. Krohova (2014). Human fetus Health classification on cardiotocographic data using random forests, Intelligent Data Analysis and its Applications, pp. 189-198.
[22] C. Buhimschi, M.B. Boyle, G.R. Saade, R.E. (1998). Garfield Uterine activity during pregnancy and labor assessed by simultaneous recordings from the myometrium and abdominal surface in the rat, Am. J. Obstet. Gynecol., 178 (4), pp. 811-822.
[23] C. Buhimschi, M.B. Boyle, R.E. Garfield, (1997). Electrical activity of the human uterus during pregnancy as recorded from the abdominal surface, Obstet. Gynecol., 90 (1), pp. 102-111.
[24] C. Buhimschi, R.E. (1996). Garfield Uterine contractility as assessed by abdominal surface recording of electromyographic activity in rats during pregnancy, Am. J. Obstet. Gynecol., 174 (2), pp. 744-753.
[25] J.S. Richman, J.R. Moorman (2000). Physiological time-series analysis using approximate entropy and sample entropy, Am. J. Physiol. - Hear. Circ. Physiol., 278 (6).
[26] E. Blinx, K.G. Brurberg, E. Reierth, L.M. Reinar, P. Oian ST (2016). Waveform analysis versus Cardiotocography alone for intrapartum fetal monitoring: a systematic review and meta-analysis of randomized trials, Acta Obstet. Gynancelogica Scand., 95 (1), pp. 16-27
[27] D Ayres-de Campos, J Bernardes, A Garrido, J Marques-de-Sá, L Pereira-Leite, (2000). SisPorto 2.0 A program for Automated Analysis of Cardiotocograms. J Matern Fetal Med. 5:311-318.
[28] Web site: https://archive.ics.uci.edu/ml/datasets/Cardiotocography#, Access date:10.7.2019.
[29] E.M. Karabulut, T. Ibrikci, (2014). Analysis of cardiotocogram data for fetal distress determination by decision tree based adaptive boosting approach, J. Comput. Commun., 2 (9), pp. 32-37.
[30] J. Spilka, G. Georgoulas, P. Karvelis, V. Chudacek (2014). Discriminating normal from ‘Abnormal’ pregnancy cases using an automated FHR evaluation, Method Artif. Intell. Methods Appl., 8445, pp. 521-531.
[31] J. Spilka, V. Chudacek, M. Koucky, L. Lhotska, M. Huptych, P. Janku, G. Georgoulas, C. Stylios, (2012). Using nonlinear features for fetal heart rate Classification, Biomed. Signal Process. Control, 7 (4), pp. 350-357.
[32] Paul Fergus, M. Selvaraj, Carl Chalmers, (2018). Machine learning ensemble modelling to classify caesarean section and vaginal delivery types using Cardiotocography traces, Computers In Biology And Medicine, 93, pp. 7-16.