Uterine Fibroid Segmentation and Measurement Based on Morphological Functions in Graphical Vision Assistant Tool

Abstract

Uterine fibroid is the most predominant problem among women of child-bearing age where the secretion of estrogen hormone plays significant role. The presence of fibroid can cause severe pain, infertility, and repeated miscarriages. Since the detection of fibroid and treatment is the crucial factor on women health especially in pregnancy, ultrasound (US) imaging is the most common modality for detecting fibroids. Because of the presence of speckle noise, the segmentation of fibroid from an US image is the tedious process. The proposed methodology has been used for automating this task by morphological functions available in graphical vision assistant tool. The modified morphological image cleaning (MMIC) algorithm for filtering and Canny edge detector have been utilized for fibroid segmentation and binary image morphological approaches adopted for analyzing the fibroid. The proposed algorithm has been developed, implemented, and validated in LabVIEW vision assistant toolbox. The outcomes of the proposed method have been evaluated and appreciated by experienced gynecologists and found that the manual intervention is eliminated on the investigation of diseased.

Classification of Uterine Fibroids in Ultrasound Images Using Deep Learning Model

“…Different ultrasound uterus image analysis methods are available in [8][9][10].S. Prabakar et al [11] defined morphological image cleaning (MMIC) algorithm to detect Uterine Fibroid. This algorithm had been developed, employed, and validated in LabVIEW vision assistant toolbox.…”