Diagnosis and Prognosis of Ultrasound Fetal Growth Analysis Using Neuro-Fuzzy Based on Genetic Algorithms

Diagnosis and Prognosis of Ultrasound Fetal Growth Analysis Using Neuro-Fuzzy Based on Genetic Algorithms

Parminder Kaur (Guru Nanak Dev University, Amritsar, India), Prabhpreet Kaur (Guru Nanak Dev University, Amritsar, India) and Gurvinder Singh (Guru Nanak Dev University, Amritsar, India)
DOI: 10.4018/978-1-7998-2742-9.ch015
OnDemand PDF Download:
No Current Special Offers


Acquisition of the standard plane is the prerequisite of biometric measurement and diagnosis during the ultrasound (US) examination. Based upon the analysis of existing algorithms for the automatic fetal development measurement, a new algorithm known as neuro-fuzzy based on genetic algorithm is developed. Firstly, the fetal ultrasound benchmark image is auto-pre-processed using normal shrink homomorphic technique. Secondly, the features are extracted using gray level co-occurrence matrix (GLCM), grey level run length matrix (GLRLM), intensity histogram (IH), and rotation invariant moments (IM). Thirdly, neuro-fuzzy using genetic approach is used to distinguish among the fetus growth as abnormal or normal. Experimental results using benchmark and live dataset demonstrate that the developed method achieves an accuracy of 97% as compared to the state-of-the-art methods in terms of parameters such as sensitivity, specificity, recall, f-measure, and precision rate.
Chapter Preview


There is a significant advancement in the analysis of ultrasound imaging and now use of modern machine learning into the medical image analysis field, this art is more challenging for the researchers (Maraci et al, 2017). Computer Aided Diagnosis (CAD) has become one of the major research subjects in medical imaging and diagnostic radiology. Ultrasound imaging modality is quite popular and most widely used modality for visualizing and studying the medical images for any disease conditions without causing any pain or discomfort to the patient (Ali et al, 2014; Jalalian et al, 2013; Cheng et al, 2010). Ultrasound imaging is widely used due to less costly and non-persistent nature as compared to other imaging modalities. The diagnosis is performed on various diseases based on image features such as the Echogenicity, Legion Shape, and Echo Texture (Kalyan et al, 2014). Accurate ultrasound-based fetal biometric measurements are important for delivery of high quality obstetrical health care. Common measurements include: the Bi-Parietal Diameter (BDP), Head Circumference (HC), Abdominal Circumference (AC), Femur Length (FL), Humerus Length (HL), and the Crown Rump Length (CRL). The American Institute of Ultrasound in Medicine (AIUM) publishes guidelines for measuring these values. These values help diagnose fetal pathology including growth restriction, microcephaly, and macrosomia. In addition, these are utilized to estimate the gestational age (GA) of the fetus (i.e., length of pregnancy in weeks and days). Accurate estimation of GA is important to determine the expected delivery date, assess the fetal size and monitor fetal growth as in Fig. 1 (Carneiro et al, 2017; Loughna et al, 2009; Hearn, 1995; Pramanik et al, 2013; Espinoza et al, 2013).

Figure 1.

Fetal Ultrasound images of (a) head, (b) femur, (c) abdomen, (d) whole fetus of age 13 weeks


The objective of this chapter is to perform Image enhancement using application of Ultrasound Fetus Development through the study of various Denoising Filtering, Feature Extraction and Machine Learning techniques. It also evaluates the quantitative denoising filtering techniques on the basis of Peak Signal-to-Noise Ratio (PSNR), Signal-to-Noise Ratio (SNR), Edge Preservation Index (EPI) and Coefficient-of-Correlation (CoC) measures. To achieve the desired objective, performance of Artificial Neural Network (ANN) based Ultrasound Fetus Development classification is assessed. Based upon outcome, hybrid Neuro-Fuzzy based Genetic approach on Ultrasound Fetus growth is proposed and compared with the existing technique (ANN) with respect to the parameters i.e. Kappa statistic, Mean absolute error, Root Mean Squared Error (RMSE), Relative absolute error, Root relative squared error, Accuracy, F-measure, Precision, Recall and Receiver Operating Characteristics (ROC) Area. 10 Fold Cross-Validation on the basis of Confusion Matrix and ROC in Classification Learner (MATLAB 2015b) has been applied to validate the results.



The ever increasing size of patient data in medical images imposes a research challenge during the scientific behavior for diagnosing, detecting and prediction of the diseases. Nowadays, the interests of

Complete Chapter List

Search this Book: