Cardiotocogram Data Classification using Random Forest based Machine Learning Algorithm

• Main Authors: Molla, M. M. Imran, Jui, Julakha Jahan, Bari, Bifta Sama, Rashid, Mamunur, Hasan, Md Jahid
• Format: Conference or Workshop Item
• Language: English; English
• Published: Springer Singapore 2020
• Subjects: QA Mathematics; TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://umpir.ump.edu.my/id/eprint/27518/1/Cardiotocogram%20Data%20Classification%20using%20Random1.pdf; http://umpir.ump.edu.my/id/eprint/27518/2/Cardiotocogram%20Data%20Classification%20using%20Random.pdf; http://umpir.ump.edu.my/id/eprint/27518/; https://doi.org/10.1007/978-981-15-5281-6_25

The Cardiotocography is the most broadly utilized technique in obstetrics practice to monitor fetal health condition. The foremost motive of monitoring is to detect the fetal hypoxia at early stage. This modality is also widely used to record fetal heart rate and uterine activity. The exact analysis of cardiotocograms is critical for further treatment. In this manner, fetal state evaluation utilizing machine learning technique using cardiotocogram data has achieved significant attention. In this paper, we implement a model based CTG data classification system utilizing a supervised Random Forest (RF) which can classify the CTG data based on its training data. As per the showed up results, the overall performance of the supervised machine learning based classification approach provided significant performance. In this study, Precision, Recall, F-Score and Rand Index has been employed as the metric to evaluate the performance. It was found that, the RF based classifier could identify normal, suspicious and pathologic condition, from the nature of CTG data with 94.8% accuracy.