Follicle Segmentation from Ovarian USG Image Using Horizontal Window Filtering and Filled Convex Hull Technique
Ultrasound imaging is the best medical imaging technology to observe and monitor the growth and physiological status of the follicles, most importantly the paramount or dominant follicle in the female’s ovary. But ultrasound images are always heavily pois
- PDF / 392,703 Bytes
- 9 Pages / 439.37 x 666.142 pts Page_size
- 85 Downloads / 197 Views
Abstract Ultrasound imaging is the best medical imaging technology to observe and monitor the growth and physiological status of the follicles, most importantly the paramount or dominant follicle in the female’s ovary. But ultrasound images are always heavily poisoned by speckle noises although it is extensively used in infertility treatment. In this paper, a segmentation technique has been developed and discussed to completely remove the speckle noises and segment different follicles from ultrasound images. The proposed segmentation technique used a 20 pixel long window and standard deviation of the USG image for smoothing and despeckling the image. Further, morphological opening followed by morphological closing operations have been applied to the image for removing the paper and salt noise. Next, segmentation of the follicles is done by finding the active contours and filled convex hull from the intermediate USG image that contains only the follicles those are bright i.e. white in color with a black background. Follicles are properly classified and detected by applying a set of relevant parameters. Finally, a comparative study has been presented between the experimental results and inferences made by the experts to validate the result towards determining the degree of accuracy of the proposed technique. Keywords Ultrasound image · Image segmentation · Active contour · Convex hull · Salt and paper noise · Image despeckling · Paramount or dominant follicle · Ovary · Morphological opening and closing
A. Mandal · M. Sarkar (B) Department of Computer Science and Application, University of North Bengal, Siliguri 734013, West Bengal, India e-mail: [email protected]; [email protected] A. Mandal e-mail: [email protected]; [email protected] D. Saha Department of Computer Science, University of Gour Banga, Malda 732103, West Bengal, India e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2021 D. Bhattacharjee et al. (eds.), Proceedings of International Conference on Frontiers in Computing and Systems, Advances in Intelligent Systems and Computing 1255, https://doi.org/10.1007/978-981-15-7834-2_52
555
556
A. Mandal et al.
1 Introduction Proper Analysis of the health and condition of developing follicles in female’s ovary is the driving force for diagnosing the female’s reproductive system. In general, this is performed by scrutinizing the Ultrasound images of ovary by the human experts. But this procedure has two major pitfalls. Firstly, it is time consuming and secondly, the accuracy level totally depends on the experts that may lead to error due to misjudgments by the experts. Today we live in the modern era of computer applications which is becoming an inseparable part of whole sphere of our life. It plays a vital role in modern medical diagnosis too, especially in medical imaging and analysis techniques. Positron Emission Tomography (PET), Ultrasound Imaging (USG), Computerized Tomography (CT) Magnetic Resonance Imaging (MRI) etc. are the widely used medical imaging techniques
Data Loading...
