ATheoretical Study on Prenatal Hydronephrosis: Image Segmentation Techniques and CNN Model

Prenatal Hydronephrosis (PNH) is caused in fetal infants identified by obstruction in the urine production, it leads to the enlargement of the kidney in fetal infants. Hydronephrosis is a common inherited defect that causes dilatation of the renal pelvis and calyces in newborns. The hydronephrosis disease grading is developed by the society for fetal urology (SFU) to get the accurate diagnosis for the decision making. This method involves the diverse population of ultrasound images of patients. Preprocessing methods in the images are used to enhance the visual quality of the image and reduce noise in the ultrasound images. This research study involves the filtering algorithms to segment the ultrasound images and it includes the convolutional neural network for classification. The severity of the hydronephrosis is measured by the extent of renal pelvis and calyceal dilation. The outcome of the study is to predict the prenatal hydronephrosis in the early stage by using the network model, the segmentation of the image shows the predictive capability of the model in distinguishing mild, moderate, and severe cases of hydronephrosis.