1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Title:	Computerized Intrapartum Cardiotocograph Pattern Features Extraction And Classification
Authors:	Shahad N. Al-Yousif (P39877)
Supervisor:	Mohd Alauddin Mohd Ali, Professor Dr.
Keywords:	Computerized Intrapartum Cardiotocograph Pattern Computerized Intrapartum Cardiotocograph Pattern Features Extraction And Classification Computerized Intrapartum Cardiotocograph Pattern Features Extraction Computerized Intrapartum Cardiotocograph Pattern Features Classification Fetal heart rate monitoring
Issue Date:	13-Sep-2013
Description:	Cardiotocograph (CTG) is a widely used non-invasive tool for assessing fetal conditions in the intrapartum and antepartum periods. Continuous monitoring of the fetal heart rate (FHR) using CTG was introduced into clinical practice in the late 1960s and since then it has been considered as an indispensable tool for fetal surveillance. CTG is a combination of the use of Doppler ultrasound (USD) for FHR determination and an external pressure transducer for recording uterine contraction (UC). Ever since its introduction in everyday clinical practices there has been a drastic reduction in child mortality. However, there are many problems related to the presence of unavoidable artifacts such as missing beats and high frequency noise. Additionally, the analysis of CTG tracings which is still being done manually by obstetricians leads to a high degree of inter-observer and intra-observer variability. Developing new methods for FHR features extraction and CTG classification is the main challenge in this research to support obstetricians in CTG analysis and reduce fetal mortality. The main objective of this research is to develop a simple reliable algorithm for FHR feature extraction and systematic CTG classification based on rules of the Royal College of Obstetricians and Gynaecologists (RCOG) guideline. Firstly, signal enhancement is implemented by compensating missing beats and removing high frequency noises. Secondly an application was developed to extract the basic features of FHR, such as: baseline, baseline variability, acceleration and deceleration based on the RCOG guideline. The implemented algorithm is based on the calculation of a virtual baseline as the algorithm reference to calculate the real baseline based on RCOG definition of baseline visual interpretation. FHR acceleration, variability and deceleration are identified accordingly. UC is the key factor considered to identify the types of decelerations whether they are early or late decelerations. A classification system was introduced to classify the CTG signal patterns into normal, suspicious or pathological CTG based on RCOG guideline. The algorithm is implemented in MATLAB source codes while the classification steps is also implemented using fuzzy logic tool box. The quality of the enhanced CTG signals after pre-processing was confirmed by five obstetricians. For the computerized interpretation of CTG signals, the results obtained for 80 different (clinical and synthetic) signals, were compared with the visual interpretation results by the five obstetricians. The obtained results show slight difference up to +/- 4 beats per minute (b.p.m.) for baseline for both synthetic and clinical data, and up to +/- 5 (b.p.m.) for variability. For the number of accelerations the difference is up to +/- 4, while the same number and type of decelerations are obtained. The fuzzy logic implementation and MATLAB codes produced the same results in classifying the CTG. Kappa statistics were implemented for the validation of classification results. The results obtained imply that the proposed methods and algorithms for CTG signal enhancement and computerized CTG interpretation are promising for use as homecare facilities for pregnant women with minimum intervention from the medical system.,PhD
Pages:	142
Call Number:	TA164 .A459 2013 3
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.